| October 10, 2019

We Tested 8 VPNs and Hotspot Shield is Fastest

VPNs are a key component for today’s savvy web user whether you’re concerned about online privacy or streaming the latest episode of an overseas baking show. But VPNs also have a reputation for destroying download speeds. We tested eight popular VPNs to see which rises to the top by keeping your connection fast.

Methodology

vpn-diagram

We tested each VPN under two circumstances: local and international. A local result means we connected the VPN to our hometown (Seattle) and then ran several tests using Speedtest.net to the closest Speedtest server. This is similar to what you might expect to see if you simply turned on the VPN with automatic settings and started browsing. Your browsing would be private, but you wouldn’t enjoy any of the geographic benefits of a VPN.

To test international results, we connected each VPN to a London VPN server and ran a Speedtest to a London Speedtest server. This simulates the experience of accessing TV shows and other media outside of your home country.

The percent difference for both local and international tests is calculated against the mean of our baseline Speedtest results when no VPN was connected. Note that while you can choose between multi-connection and single-connection testing on most of our Speedtest platforms, the results outlined below used multi-connection.

All VPNs slowed our local connections

Ookla_Difference-in-Speed-When-Testing-VPN-to-Local-Server
The greater the percent difference from the baseline, the slower your connection will be. Hotspot Shield had the least impact to download speed in local testing with a decrease of 51.9%. IPVanish was second best — decreasing download speed by 66.7%. Private Internet Access was third at -70.9%, and TunnelBear had the slowest speeds with -98.3%. We should note that while IPVanish, Hotspot Shield and NordVPN have business relationships with Ookla, our results are independent of these relationships.

Hotspot Shield was actually faster than the baseline on our international tests

Ookla_Difference-in-Speed-When-Testing-VPN-to-Remote-Server-2
Our international VPN test showed Hotspot Shield in first place with a 26.2% increase in download speed when testing between Seattle and London. We spoke to our engineers to find out what allows Hotspot Shield’s technology to improve speeds internationally. Hotspot Shield funnels network TCP traffic through a transparent TCP proxy, from a location on the internet close to the device, to the remote destination. This proxy efficiently manages congestion that typically occurs at the last mile which allows for an increase in speeds versus standard routing.

All other VPNs showed the expected decrease in download speeds when testing internationally. ExpressVPN was the second fastest VPN for international connections with a 42.5% decrease in download speed. Avira was third with an 85.2% decrease. IPVanish and NordVPN followed closely behind, coming in within 1% of each other. TunnelBear, again, had the slowest speeds with a 95.3% decrease.

Now that you know which VPN is fastest, take a Speedtest next time you’re connected to a VPN to see how protecting your online privacy is impacting your speeds.

Ookla retains ownership of this article including all of the intellectual property rights, data, content graphs and analysis. This article may not be quoted, reproduced, distributed or published for any commercial purpose without prior consent. Members of the press and others using the findings in this article for non-commercial purposes are welcome to publicly share and link to report information with attribution to Ookla.

| January 31, 2018

Which Mobile Carrier Won Sunday’s Big Football Game?

Football’s biggest game is this weekend and we’re excited to hear if all the reported mobile and Wi-Fi investments at U.S. Bank Stadium are up to the challenge on Sunday. We’ll follow up on that next week, but in the meantime we’re limbering up by checking out mobile internet speeds at the Eagles’ and Patriots’ home stadiums over the season. We’re also offering a preview of the performance we’ve seen at U.S. Bank Stadium so far this season.

We looked at stadium speeds for August 1, 2017 through January 30, 2018 and compared speeds by carrier using Speed Score — a combined measure of each provider’s download and upload speed at the 10th, 50th and 90th percentiles — on modern devices (i.e. not your grandfather’s flip phone).

Lincoln Financial Field, home of the Eagles

If we were handicapping teams based on mobile speeds at their home fields, the Eagles would be in sore shape for Sunday. Although Verizon had the best speeds of the season, the carrier’s Speed Score at Lincoln Financial Field was 63.6% lower than it was at Gillette Stadium. Still, Verizon’s Speed Score was 59.2% higher than AT&T’s, 121.3% higher than Sprint’s and 220.7% higher than T-Mobile’s.


Speedtest Cellular Internet Results

Lincoln Financial Field | 2017 Season
Carrier speeds based on Speed Score for modern devices

Carrier Speed Score
Verizon Wireless 27.90
AT&T 17.52
Sprint 12.61
T-Mobile 8.70

Gillette Stadium, where the Patriots go to win

The Patriots won’t need to stage one of their famous comebacks to claim a win for mobile internet speeds on their home turf. Looking at speeds at Gillette Stadium, Verizon wins again with the fastest Speed Score of them all. Add to that every major carrier had faster speeds here than they did in Philadelphia. Verizon was 174.4% faster in Foxborough than in Philly, AT&T was 25.2% faster, Sprint jumped 143.9% and T-Mobile saw a 173.0% increase in Speed Score when comparing their performance at Gillette Stadium with that at Lincoln Financial Field.


Speedtest Cellular Internet Results

Gillette Stadium | 2017 Season
Carrier speeds based on Speed Score for modern devices

Carrier Speed Score
Verizon Wireless 76.56
Sprint 30.76
T-Mobile 23.75
AT&T 21.93

U.S. Bank Stadium, the place to be on Sunday

The U.S. Bank Stadium in Minneapolis was designed with big events like Sunday’s football game in mind from the capacity, to the readerboards to the app. But what we’re really interested in is the tech. With Wi-Fi access points literally built into the railings, U.S. Bank Stadium should be the ideal place to live-stream any game or event. As for cellular, there’s a distributed antenna system that should keep fans connected.

Season internet speeds play-by-play

Examining internet performance at U.S. Bank Stadium for all home games during the regular stadium, the impressive speeds bode well for the big day.

Team cellular vs. team Wi-Fi

Cellular speeds at U.S. Bank Stadium far exceeded the U.S. average throughout the 2017 football season. On the date with the slowest speeds, September 24, downloads were still 62.6% faster than the average for the U.S. during that month and 5.9% faster than Minneapolis’s average download speed for September 2017.


Speedtest Cellular and Mobile Wi-Fi Results

U.S. Bank Stadium | 2017 Season
Stadium speeds based on mean speeds for all results

Day Cell – Mean Download (Mbps) Cell – Mean Upload (Mbps) Wi-Fi – Mean Download (Mbps) Wi-Fi – Mean Upload (Mbps)
Sep 11, 2017 59.61 23.14 27.87 28.45
Sep 24, 2017 38.99 18.30 21.65 25.75
Oct 1, 2017 41.98 21.35 21.82 28.07
Oct 15, 2017 62.54 23.54 29.37 25.92
Oct 22, 2017 53.05 26.83 30.01 29.11
Nov 19, 2017 69.02 24.97 22.03 25.38
Dec 17, 2017 59.25 22.05 18.84 17.65
Dec 31, 2017 63.15 22.88 14.10 16.41

Meanwhile, the Wi-Fi might be free at U.S. Bank Stadium and the average speed does exceed that of the free Wi-Fi we’ve seen at many North American airports, including LAX and JFK, but it’s still not as fast as we’d hoped. Especially given all the hype. Free Wi-Fi is a great resource to have, but in most cases you’ll get faster speeds using your mobile carrier’s network on game day.

Carrier performance

T-Mobile had the fastest Speed Score at U.S. Bank Stadium. On average, T-Mobile was 18.6% faster than Verizon, 65.1% faster than AT&T and 107.4% faster than Sprint.


Speedtest Cellular Internet Results

U.S. Bank Stadium | 2017 Season
Carrier speeds based on Speed Score for modern devices

Carrier Speed Score
T-Mobile 87.44
Verizon Wireless 73.74
AT&T 52.96
Sprint 42.15

What to expect during the big game

Our technical evangelist, Milan Milanović, dug a little deeper into some upgrades mobile carriers have been working on to make sure their networks are game-ready. Here’s what he found:

  • AT&T’s upgrades include deploying over 800 antennas throughout the stadium, providing a more than 200% increase in capacity. Additional legacy spectrum assets have been refarmed (Band 5) and coupled with advanced LTE technologies. AT&T is expected to keep up with the increased traffic demand.
  • Verizon has added 48% more antenna nodes at the stadium, which are largely invisible (tucked under the seats, handrails, etc.). In addition to over 50 MHz of deployed capacity, Verizon is activating Higher Order MIMO and Higher Order Modulation together with three-channel carrier aggregation. You could say Verizon has their game face on.
  • Sprint’s network within the stadium also relies on 800 nodes and a distributed antenna system (DAS) powered by small cells. This will deliver 40 MHz of TDD capacity in the 2.5 GHz frequency band. Sprint will rely on two-channel carrier aggregation in the downlink to provide required downlink capacity. It will be interesting to see if this network configuration will be sufficient to support the inevitable spike in upload traffic, especially during halftime.
  • T-Mobile has increased the node capacity within the stadium by a factor of 30 in addition to the already deployed 4×4 MIMO, 256 QAM and three-channel carrier aggregation LTE techniques. They have also allocated additional spectrum assets to LTE in order to provide 45 MHz of downlink capacity, upgraded backhaul at and around the event and centralized radio access technology by the way of uplink CoMP (Coordinated Multipoint). This centralized radio access technology is designed to improve uplink data rates and network efficiency, with the added benefit of extending a smartphone’s battery life.

His assessment for fans on the big day? Bring your best devices for the best performance (for you and others). Users with smartphones equipped with four receive antennas will be able to experience the fastest speeds possible. In the process, they’ll also contribute to higher network efficiency by virtue of more quickly completing tasks that improve spectrum resource utilization and therefore the overall capacity of the network. Sunday’s event will also be a test of which operator has most aggressively seeded these highly efficient handsets into the hands of consumers.

Game day stats

You’ve probably seen the wide swing in reported speeds from the big game, and, now that Eagles fans have almost recovered from Sunday’s celebration, we’re here to set the record straight.

Cellular speeds

Looking at speeds at U.S. Bank Stadium between 3:30 p.m. and 8 p.m. on Sunday, we found that the mean cellular download speed at the stadium was faster than we’d seen at any game, all season long. Not a lot faster, but, given the 7.2 terabytes of data used during Sunday’s game, we’re impressed.


Speedtest Cellular Internet Results

U.S. Bank Stadium | During the Big Game
Stadium speeds based on mean speeds for all results

Day Cell – Mean Download (Mbps) Cell – Mean Upload (Mbps)
Feb 4, 2018 69.09 15.93

Which carrier won?


Speedtest Cellular Internet Results

U.S. Bank Stadium | During the Big Game
Carrier speeds based on Speed Score for modern devices

Carrier Speed Score
T-Mobile 114.72
AT&T 58.54
Verizon Wireless 50.66
Sprint 33.24

Comparing carriers based on Speed Score between 3:30 p.m. and 8 p.m. on Sunday, T-Mobile not only came out on top, but they trounced their season average, with a 31.2% jump in speeds. T-Mobile’s game-winning Speed Score was also 96.0% faster than runner-up AT&T who came in second and saw a 10.5% increase in speeds on game day.

Despite significant efforts on the technology side, the other two major carriers saw declines in Speed Scores on the big day. Verizon Wireless came in third and saw a 31.3% decrease from their season average while Sprint’s fourth-place finish was 21.1% slower than their all-season number.

How ‘bout that Wi-Fi?

We could not find any Speedtest results on “#USBANKSTADIUM”, the stadium’s reported Wi-Fi SSID, which leads us to believe that the powers that be switched the SSID to “#SBFAN” for the big event. If this is the case, the game day Wi-Fi download speed at U.S. Bank Stadium was in the low to mid range of what we saw during the rest of the season.


Speedtest Mobile Wi-Fi Results

U.S. Bank Stadium | During the Big Game
Stadium speeds based on mean speeds for all results

Day Wi-Fi – Mean Download (Mbps) Wi-Fi – Mean Upload (Mbps)
Feb 4, 2018 20.71 23.36

Was the big game worth the big investment? Likely. Huge crowds hungry to share the experience on social media can clog up networks and destroy cellular speeds. We’ll be interested to see how this year’s performance changes carriers’ playbooks for 2019.!

Ookla retains ownership of this article including all of the intellectual property rights, data, content graphs and analysis. This article may not be quoted, reproduced, distributed or published for any commercial purpose without prior consent. Members of the press and others using the findings in this article for non-commercial purposes are welcome to publicly share and link to report information with attribution to Ookla.

| February 3, 2020

Verizon Rocks Mobile Speeds at the Big Game in Miami

More than 65,000 people watched the Kansas City Chiefs beat the San Francisco 49ers live at the Hard Rock Stadium in Miami yesterday. All those fans on their phones sharing the halftime experience with folks back home used massive amounts of data as they livestreamed and posted to social media. Every year this event presents a major challenge to mobile operators who this year added 5G to the mix. We’re here to report on which operator had the fastest speeds and the lowest latency on game day and what indoor coverage looked like in the Miami area.

Stadium-Heatmap

Mobile operators had their game face on

This is no ordinary sporting event, and operators have been working for many months on their playbooks for providing the best possible performance at Hard Rock Stadium. Highlights include:

Verizon Wireless’s download speeds beat out competitors’

We compared the big four U.S. mobile operators from two hours prior to kick-off to 30 minutes after the game ended to see who won. Here’s what we saw:

Overall Cellular Performance at Hard Rock Stadium
Speedtest® Data | February 2, 2020
Operator Mean Download Speed (Mbps) Mean Upload Speed (Mbps) Latency (ms)
Verizon Wireless 297.18 9.81 59
T-Mobile 121.93 30.34 34
Sprint 114.93 6.20 51
AT&T 103.68 8.38 46

Sprint’s home team might have won the game, but Verizon triumphed when it came to mobile download speeds, coming in 143.7% faster than second-place T-Mobile when considering Speedtest® results over all technologies. Sprint had the third fastest mean download speed in this category and AT&T came in fourth.

Focusing on their 5G game really helped Verizon take the day when considering overall speeds as T-Mobile had the fastest mean download speed (66.35 Mbps) on LTE. Sprint was second on LTE with a mean download speed of 56.16 Mbps, AT&T third at 39.18 Mbps and Verizon fourth at 30.67 Mbps. We break out 5G speeds for each operator below.

T-Mobile’s mean upload speed over all technologies was far better than competitors’. Upload speed is especially important at big events like this as fans try to share their game day experience with those not in the stadium.

T-Mobile also had the lowest latency, coming in 26.0% faster than second-place AT&T. Sprint was third for latency and Verizon fourth.

For comparison, the mean download speed over mobile in the U.S. in January 2020 was 41.23 Mbps, upload was 10.55 Mbps and latency was 46 ms.

5G for the win

5G is the biggest story in mobile these days, and all four operators came prepared to deliver their best game to customers with 5G-capable phones and compete for bragging rights during this high profile event.

5G Performance at Hard Rock Stadium
Speedtest® Data | February 2, 2020
Operator 5G Download Speed (Mbps) 5G Upload Speed (Mbps) 5G Latency (ms)
Verizon Wireless 646.17 9.86 91
T-Mobile 348.33 28.89 32
Sprint 225.78 15.39 15

Verizon easily beat T-Mobile and Sprint when it came to mean download speed over 5G during the big game. Verizon came in last, however, for both mean upload speed over 5G and latency. T-Mobile showed the fastest mean upload speed over 5G and Sprint had the best latency on 5G. While we did see 5G Speedtest results for AT&T during the game, there were fewer than 10, the minimum threshold we set for this event.

Hard Rock Stadium Wi-Fi was a viable option

In 2019, fans at Mercedes-Benz Stadium in Atlanta used over 24 TB of data on the stadium Wi-Fi network on game day with an average Wi-Fi download speed of 30.98 Mbps. To get a baseline on the Miami experience, we looked at Wi-Fi performance during the Bengals v. Dolphins matchup on December 22, 2019.

Wi-Fi Performance at Hard Rock Stadium
Speedtest® Data | February 2, 2020
Mean Download Speed (Mbps)Mean Upload Speed (Mbps)Mean Latency (ms)

Stadium Wi-Fi – December 22, 2019 56.48 62.64 7
Stadium Wi-Fi – Big Game 37.43 46.55 9
Verizon Wi-Fi – Big Game 36.81 40.91 7

Wi-Fi at the stadium did show some game day stresses yesterday, with a mean download speed 33.7% slower than the December 22 game. Mean upload speed dropped 25.7% and latency was up 28.6%.

Verizon also provided Wi-Fi for their customers during the big game yesterday, and the mean download speed was comparable to that on the stadium’s SSID. Mean upload speed on Verizon’s SSID was 12.1% slower than on the stadium’s, but Verizon’s Wi-Fi latency was also lower, showing a 22.2% improvement over stadium Wi-Fi.

It’s worth noting that mean upload speed in all cases was faster than that on download. This is impressive and helpful to fans trying to livestream their experience for friends back home.

Indoor mobile coverage in Miami

Stadium speed is very important, however most attendees will spend the bulk of their time in hotels and at tourist attractions in and around Miami. We used Cell Analytics to assess indoor coverage at 9,128 buildings in the Miami metro area during Q3-Q4 2019 to see who had good indoor coverage (signal strength of greater than -115 dBm) in the most buildings.

Indoor Mobile Coverage in Miami Metro Area
Cell AnalyticsTM Data | Q3-Q4 2019
Operator % of Buildings with Good Coverage % of Buildings with Best Coverage
Verizon 97.7% 29.1%
T-Mobile 97.2% 36.9%
AT&T 94.2% 22.7%
Sprint 92.5% 11.2%

Verizon showed good coverage in more buildings than any other operator at 97.7%, followed by T-Mobile at 97.2%, AT&T at 94.2% and Sprint at 92.5%.

We also examined who had the best coverage in each building and found that T-Mobile provided the strongest coverage in 36.9% of buildings analyzed. Verizon had the strongest coverage in 29.1% of buildings analyzed, AT&T was strongest in 22.7%, and finally Sprint was strongest in 11.2%.

Learn how Ookla® can help you determine if your network is prepared for the massive crowds that accompany a marquee event and analyze how your network performs both indoors and out, down to the building level.

Ookla retains ownership of this article including all of the intellectual property rights, data, content graphs and analysis. This article may not be quoted, reproduced, distributed or published for any commercial purpose without prior consent. Members of the press and others using the findings in this article for non-commercial purposes are welcome to publicly share and link to report information with attribution to Ookla.

| March 14, 2019

Ditch the Lag: Cities with Great Gaming Culture and Low Ping

Yes, you can game from anywhere with an internet connection. But if you’re at all competitive, it’s nice to play from somewhere with low ping and fast internet speeds. Plus when you need to leave the house, it’s extra nice to know you’re also surrounded by gamer culture. We’ve examined February 2019 Speedtest results in 35 cities that are known for their esports events, gaming conferences, game companies and more to find out who has the advantage and ranked them based on their ping.

The top contenders

Eleven_Gaming_Cities_0219

First place Bucharest, Romania is home to super-low ping, a lightning fast download speed and a thriving gaming culture. From Bucharest Gaming Week (which includes the CS:GO Southeast Europe Championship and the FIFA National Tournament) to their numerous local game studios, Bucharest is a great place to be a gamer whether you’re online or out and about.

The next five gaming cities with the lowest pings are all in Asia. Hangzhou, China comes in second overall with a fast ping and world-class download speeds. This city is so devoted to its gamers that it opened a $280 million gaming “city” in 2018 and plans 14 new esports arenas before 2022. Coming in third, Chengdu, China has an equally low ping to our first two contenders and serves as one of two host locations in China for the Global Mobile Game Confederation (GMGC). Both Hangzhou and Chengdu are also franchise holders in the Overwatch League, giving local gaming fans something to cheer about. Fourth place Singapore, host of the 5th Annual GameStart Convention in October 2018, had only a slightly slower ping than the first four cities and the fastest download speed of any of the cities we considered.

South Korea is home to the fifth and sixth best cities for gamers. A satellite city of Seoul, Seongnam-si boasts the Pangyo Techno Valley (a.k.a. the Silicon Valley of Korea) and numerous game development companies. Perfect for a city with a 9 ms ping. Though Incheon’s ping was a little slower at 12 ms, gamers there can console themselves with the city’s gamer cred — the 2018 League of Legends World Championship was held in Incheon’s Munhak Stadium.

Coming in at number seven, Budapest, Hungary is an emerging game city, having hosted its first big esports event (the V4 Future Sports Festival) in 2018, but a 12 ms ping makes them a strong contender. More established Malmö, Sweden is number eight with a slightly slower average download speed but the city is headquarters to Massive Entertainment, creators of Tom Clancy’s The Division series, Far Cry 3, Assassin’s Creed: Revelations and many more.

Vancouver, Canada, North America’s only qualifier for the top gaming cities list, comes in at number nine with a 12 ms ping and many gaming companies including the Canadian arms of Nintendo of Canada and EA (Electronic Arts). We included both Shanghai, China and Moscow, Russia on the top gamer cities list as both had a 12 ms ping as well, though the internet speeds in Shanghai are superior. Shanghai will also host the International Dota 2 in 2019 while Moscow is known for Epicenter.

The rest of the pack

Notably absent from the list above is most of the western hemisphere. Cities in North America were held back by their high pings. Cities in South America suffered from high pings and also slow internet speeds — something that esports leagues have complained is a barrier to investment.

Our full list of gaming cities provides wider geographical representation, even if the internet performance is not always as stellar. You’ll find Los Angeles in 27th place, behind Seattle, Boston and Las Vegas. And São Paulo, Brazil has the best showing in Latin America at 23rd.

Internet Performance in 35 Cities with a Gaming Culture
Speedtest Results | February 2019
City Ping (ms) Mean Download (Mbps) Mean Upload (Mbps)
Bucharest, Romania 8 172.13 126.57
Hangzhou, China 8 125.93 29.54
Chengdu, China 8 101.92 33.80
Singapore 9 196.43 200.08
Seongnam-si, South Korea 9 155.25 114.83
Incheon, South Korea 12 139.84 102.91
Budapest, Hungary 12 132.72 54.46
Malmö, Sweden 12 126.28 105.67
Vancouver, Canada 12 117.55 50.23
Shanghai, China 12 75.14 30.06
Moscow, Russia 12 64.56 63.59
Oslo, Norway 13 115.46 69.03
Hong Kong, Hong Kong (SAR) 14 167.59 161.14
Zürich, Switzerland 14 144.36 109.39
Seattle, United States 15 138.50 79.88
Stockholm, Sweden 15 134.16 93.83
Auckland, New Zealand 15 92.05 53.30
Toronto, Canada 16 134.75 67.42
Boston, United States 17 152.42 60.87
Las Vegas, United States 17 141.69 41.22
Chennai, India 17 48.40 42.93
Cologne, Germany 18 63.77 18.36
São Paulo, Brazil 18 46.43 21.57
Jakarta, Indonesia 18 17.88 10.21
Mumbai, India 19 23.40 19.26
Paris, France 20 161.04 93.68
Los Angeles, United States 20 121.00 23.57
London, United Kingdom 20 63.58 23.18
Rio de Janeiro, Brazil 20 36.50 13.33
Buenos Aires, Argentina 21 34.31 6.40
Katowice, Poland 22 83.99 20.91
Mexico City, Mexico 25 37.66 15.39
Sydney, Australia 25 34.20 9.61
Santiago, Chile 26 56.13 18.49
Tokyo, Japan 28 99.24 101.90

Of course, die-hard gamers will know that a low ping in your city won’t necessarily save you if you’re playing on a distant server.

What’s the ping like in your city? Take a Speedtest and see if your connection is hurting your gameplay.

Ookla retains ownership of this article including all of the intellectual property rights, data, content graphs and analysis. This article may not be quoted, reproduced, distributed or published for any commercial purpose without prior consent. Members of the press and others using the findings in this article for non-commercial purposes are welcome to publicly share and link to report information with attribution to Ookla.

| May 18, 2022

Four U.S. Airports Top Our List of Fastest Free Airport Wi-Fi

The COVID-19 pandemic upended a lot of travel plans and put our series on airport Wi-Fi on hold. We’re excited to get back in the swing of things in time for the summer travel season, and we have fresh data for you on Wi-Fi speeds at some of the busiest airports in the world during Q1 2022. Our analysis focuses on Wi-Fi over mobile connections on free Wi-Fi provided by the individual airports and Wi-Fi at selected airport lounges. The good news is that every airport surveyed met the recommended speed for streaming on mobile. However, there was a large divide between the fastest airports on the list and the slowest.

Mobile Internet Speeds Over Free Wi-Fi at Select Airports

Speedtest Intelligence® showed four airports in the United States at the top of the list for free airport Wi-Fi. San Francisco International Airport showed a median download speed of 176.25 Mbps during Q1 2022, Seattle-Tacoma International Airport 171.01 Mbps, Dallas/Fort Worth International Airport 113.64 Mbps, and Chicago O’Hare International Airport 85.49 Mbps. Dubai International Airport, Hartsfield–Jackson Atlanta International Airport, Amsterdam Airport Schiphol, and Los Angeles International Airport followed. All of these airports are international hubs that passengers from around the world pass through on their way to all kinds of destinations. Flyers waiting for connecting planes at these airports should have no trouble with internet speeds. In case of video calls, upload speeds are even faster than downloads at all of these airports, and San Francisco and SeaTac had the fastest uploads on the list. There is a wide gap in median speeds between the free airport Wi-Fi at Los Angeles International Airport and the rest of the airports on our list.

Denver International has their Wi-Fi split between two SSIDs that serve different spectrum bands, both of which showed median download speeds between 44 and 46 Mbps. Connecting to the 5 GHz “- DEN Airport Free WiFi” will get you a faster median upload speed than “DEN Airport Free WiFi 2.4.” Charles de Gaulle also has two different SSIDs for their free airport Wi-Fi though travelers can select either, “*WIFI-AIRPORT” had slightly faster median speeds than “WIFI-AIRPORT-STANDARD.”

Turkey’s Istanbul Airport was next on the list with a median download speed of 31.08 Mbps and a median upload speed of 23.80 Mbps. Spain’s Madrid Barajas Airport was the top of the bottom of our list with a median download speed of 19.76 Mbps over free airport Wi-Fi.

Four of the bottom five airports on our list were in China. Shanghai Hongqiao International Airport was faster than the other Chinese airports on the list with a median download speed of 14.69 Mbps. Shenzhen Bao’an International Airport and Beijing Capital International Airport were next on our list and showed nearly identical median download and upload speeds over free airport Wi-Fi. Chengdu Shuangliu International Airport showed slightly slower download and upload speeds than both. Mexico City International Airport in Mexico had the slowest median download and upload speeds over free airport Wi-Fi of any airport on this list.

Wi-Fi in airport lounges is often faster than free airport Wi-Fi

Free Wi-Fi is very nice to have for catching up on your life back home or reliving the glory of your vacation pics as you upload them to your social media accounts, but if you’re looking for the fastest Wi-Fi in an airport, you may want to look into other options. In many cases we found that airport lounges had faster Wi-Fi, like the United Club in Chicago and San Francisco which boasted median download speeds of 246.17 Mbps and 244.37 Mbps, respectively, during Q1 2022. The fastest club Wi-Fi download speed at LAX was at the Alaska Lounge (238.59 Mbps).

Outside the U.S., Wi-Fi speeds at airport lounges ranged dramatically. The following lounges all showed faster median download speeds than the free Wi-Fi at their respective airports during Q1 2022: the Plaza Premium Lounge in Dubai (148.96 Mbps), the Grand Lounge Elite in Mexico City (125.12 Mbps), Privium at Amsterdam’s Schiphol (121.05 Mbps), the Emirates Lounge at Paris’ Charles de Gaulle (93.31 Mbps), and the Sala VIP Puerta de Alcalá in Madrid (50.28 Mbps).

This stronger performance at airport lounges should be expected because it’s much easier to configure Wi-Fi over a small area like a single airport lounge than it is to serve an entire airport with multiple terminals. Patrons of airport lounges are also paying for the privilege, whether by the day or the year, so their expectations are higher. We did not always find this to be the case, however, so take a Speedtest® if you feel like you are not getting what you are paying for.

We were glad to see that travelers at all these airports have fast enough Wi-Fi to stream video on mobile and that there are other options available for those who need faster internet. Airports have had a lot to focus on in the last couple of years and Wi-Fi was not at the top of the list. We hope that will change as passenger volumes return to normal. If you’re traveling this summer, take a Speedtest at the airport to see how your experience compares.

Ookla retains ownership of this article including all of the intellectual property rights, data, content graphs and analysis. This article may not be quoted, reproduced, distributed or published for any commercial purpose without prior consent. Members of the press and others using the findings in this article for non-commercial purposes are welcome to publicly share and link to report information with attribution to Ookla.

| August 26, 2021

How to Identify and Resolve Network Issues in Real-Time [Webinar]


Drive testing has historically been a time-consuming, expensive and labor-intensive effort — but it doesn’t have to be. With recent advances in crowdsourced network insights, revolutionary new live testing capabilities and real-time analytics, some of the leading operators in the U.S. have drastically cut the time and budget they spend finding and fixing network issues.

The upcoming Ookla® webinar on September 9 will show how Network Optimization and RAN teams can combine crowdsourced network data with surgical drive and walk testing to make immediate network improvements in the areas that matter most to consumers. Read on to learn how mobile network operators can save countless hours and dollars with this new approach to live testing.


Use Crowdsourced network data to pinpoint areas with high user counts and poor signal or throughput

It’s impossible to drive test every street and walk test every building — but crowdsourced network data allows you to see real-world connectivity where it matters most to customers. Powered by hundreds of millions of daily performance, coverage and signal measurements from Speedtest®, Ookla Cell Analytics™ provides unparalleled intelligence about wireless service quality, RF measurements, data use, indoor vs. outdoor performance, cell site locations and much more.

By looking at user density, mobile operators can understand where the highest volume of customers are impacted by poor signal or throughput. After you’ve prioritized the areas where network improvements will have the most impact on subscribers, you can send your field testing team to conduct more targeted analysis on-site.

With Cell Analytics, we can scan wide regions to identify and prioritize problem areas. In this case, we looked at a wireless network in Las Vegas and quickly saw that the Paris Las Vegas Hotel & Casino has lots of users, but the network has very poor quality there (LTE RSRQ and SNR) and also low data speeds, despite having good coverage throughout the property.

Mobile-User-Density-in-Las-Vegas

The poor quality extends from the ground floor through the upper floors of the building. We can also see that the operator has a cell site on the property.

Indoor-LTE-RSRQ

From the crowdsourced data in Cell Analytics, we can also see that the 1900 and 2100 MHz bands are most heavily used throughout the property.

Most-Frequent-Band

eNodeB 80024 and 80155 are serving most frequently on the property, although some others are also seen.

LTE-Most-Frequent-Cell

LTE-Most-Frequent-Cell-2

Armed with this knowledge, the mobile network operator can send someone to conduct a quick walk test of the building.

Perform targeted drive and walk tests for problem areas with real-time analytics

In the past, drive and walk testing could be prohibitively expensive for smaller network operators. For example, let’s say you’re an operator with 40,000 sites, and a tester spends an average of five hours per site at $50/hr. Testing every site in your network would cost upwards of $10 million per year. Ookla Wind™ (Wireless Intelligence On Demand) offers a revolutionary approach to drive and walk testing with affordable devices, real-time analytics and no time spent on post-processing.

As we saw above, Ookla’s Cell Analytics allows us to identify the “symptoms” of poor network performance. Specific indoor areas in the main lobby and casino floor inside the Paris Las Vegas Hotel & Casino showed poor signal-to-interference-and-noise-ratio (SINR) and low throughput performance. Upon completing a surgical walk-test of the same areas using the Ookla Wind handset based network measurement platform, the operator was able to diagnose the issue.

Below we can see it was clear that the indoor area lacked any 5G connectivity. This, coupled with no carrier aggregation, low MIMO utilization and lower modulation scheme due to poor SINR, all contribute to poor throughput.

real-time-walk-test-view-2

Given the large number of handover attempts in our walk-test, the recommended next step for the operator is to address the pilot and reference signal pollution in the area and to establish clear dominance to improve the network performance. By utilizing Ookla Wind, a remote engineer could analyze the data in real-time and make these adjustments to the network while the tester is still on site. This should save you hours or days that would otherwise be spent waiting for post-processing to happen.

Processed-Report

The webinar on Thursday, September 9 at 9 a.m. PDT (12 p.m. EDT / 4 p.m. GMT) will show you how to combine crowdsourced network data with surgical drive and walk testing to make immediate network improvements. Don’t miss it. A recording will be provided for registrants who can’t tune in to the live presentation.


Ookla retains ownership of this article including all of the intellectual property rights, data, content graphs and analysis. This article may not be quoted, reproduced, distributed or published for any commercial purpose without prior consent. Members of the press and others using the findings in this article for non-commercial purposes are welcome to publicly share and link to report information with attribution to Ookla.

| March 11, 2020

Three Ways to Improve Your Network Strategy with Crowdsourced Data (White Paper)

Both wireless network operators and infrastructure companies face challenges when determining where to make investments or improvements. Wireless network operators must properly configure cell sites to minimize interference, determine where networks are performing poorly and rapidly identify what network changes to implement to improve service in those locations. Similarly, infrastructure companies, such as tower and distributed antenna system (DAS) builders, must be able to determine areas and buildings with lots of users and poor service on multiple networks, then use this information to identify the best places to build new towers or “neutral host” indoor systems to lease to as many operators as possible.

To make these decisions, most providers currently rely on expensive drive testing, performance counter data that doesn’t offer location accuracy and outdated information on population density. In this new white paper, we share how providers can use real-world network performance and quality measurements from Ookla® to inform their network strategy. Download the full white paper here.

A better way to collect network performance, coverage and signal data

Recent advances in collecting and analyzing crowdsourced measurements can help inform network optimization decisions. Ookla Cell Analytics™ surfaces real-world data on wireless service quality, RF measurements, data usage, user density and other key metrics. Cell Analytics uses the results from 10+ million daily consumer-initiated tests taken on Speedtest® and hundreds of millions of coverage scans on Speedtest Android every day to provide accurate location information, both indoors and out, down to the individual building level.

Performance-Counter_Crowdsourced-RSRP-Measurement-Views-1

The above image compares typical performance counter metrics to the detailed view in Cell Analytics of LTE signal level (RSRP) for one network operator in Philadelphia.

Locating competitor cell sites and opportunities for new cell site deployment

Using the Site Finder feature in Cell Analytics, mobile network operators and infrastructure providers can accurately estimate the locations of cell sites for all operators. In addition to competitive benchmarking, this also allows providers to discover areas with poor coverage or quality — and no existing cell sites — to identify deployment opportunities.

Houston_Poor-Coverage_High-User-Density
The above image shows individual buildings that are scored on the count of indoor users and indoor service quality on all networks. Buildings with a high score (indicated in red) may be good opportunities for network operators to increase capacity and for infrastructure providers to build new rooftop assets or neutral host systems.

In this white paper, we share how network engineering teams can use Cell Analytics data to discover the locations that need easy-to-make network improvements, find opportunities to add new cell sites and prioritize optimization efforts based on user density and coverage data from multiple networks.

Looking at real-world data from Italy, Brazil and the United States, download the full white paper to discover how operators can make immediate, data-driven improvements to their networks — without the overhead and limitations of drive testing.

Ookla retains ownership of this article including all of the intellectual property rights, data, content graphs and analysis. This article may not be quoted, reproduced, distributed or published for any commercial purpose without prior consent. Members of the press and others using the findings in this article for non-commercial purposes are welcome to publicly share and link to report information with attribution to Ookla.

| May 13, 2020

Indoor Coverage is a Public Safety Priority

Connecting people with emergency services reached a pinnacle of simplicity when 911 was rolled out as an emergency number across the United States. With one number, people in distress could get the help they needed dispatched as soon as possible. As increasing numbers of households have cut the cord on their traditional landline telephones, 80% of 911 calls are now placed via cell phone. Emergency calls need to connect 100% of the time, which makes mobile coverage, first and foremost, a public safety issue.

Understanding which buildings fall short of providing adequate service can assist local governments in working with building owners and mobile operators to make needed improvements. This falls into two broad categories: First Responder Push to Talk systems, and Commercial Mobile Services used by both Public Safety Agencies and the general public.

Poor indoor coverage impacts public safety

If someone is in distress and unable to place an outgoing call, first responders will not be aware there is an emergency that requires their response. For this reason, the Safer Buildings Coalition defines three pillars of in-building safety communications:

  • Mobile 911 Calls Must Get Out with Location Accuracy
  • Mobile Mass Notifications Must Get In
  • First Responder Communications Must Work

If a building cannot deliver these basic characteristics, the environment puts the occupants and the property itself at risk.

Determining a precise location can be a significant challenge if the device does not have an unobstructed view of the sky. As more GPS satellites can “see” the device, the more accurate the location the system can provide. Work is underway by industry leaders and public safety agencies to improve indoor location, but since it is a complex issue unto itself, this article will focus solely on indoor wireless network coverage.

Why indoor coverage is challenging

Anyone who’s ever tried to place a call from an elevator is not surprised that indoor coverage can be much worse than outdoor coverage. And the deeper into a building you go, the worse the signal typically gets. Penetrating walls is difficult for a cellular signal, though some of the spectrum blocks that mobile companies have licenced are better for this task than others. Low band (longer wavelengths) spectrum tends to be much better at penetrating concrete and brick than high band (shorter wavelength) spectrum.

Low-e glass can inhibit signals

Another factor in poor indoor signal strength is often windows. The introduction of low-e glass has provided huge energy savings for building owners and is positive for the environment. However, the unintended effect is a negative impact on wireless communications.
SBC-Low-E-Glass_Illustration-1

How glass compares to other building materials in shielding the interior from wireless signals depends upon the type of glass. The chart below offers some surprising comparisons. The attenuation column represents the reduction in the amplitude of the signal. For this example we use 900 MHz, a common low-band spectrum used throughout most of the world and considered to be better at building penetration than higher band frequencies.
building_material_effect_cellular_signals_0520

The more energy efficient the glass, the more the signal level is reduced. Consider that for each 3 dB attenuation (loss), the signal strength is reduced by half. A 6 dB attenuation means a 75% loss in signal; at 9 dB, an 87.5% loss. As this reduction is exponential, the double glazing low-e glass, near the bottom of the chart, represents a signal reduction of 99.9%.

It’s not uncommon to see someone who is struggling to maintain a call walk toward the window in an attempt to improve their reception. If a building has installed energy efficient glass, most of the available signal may well be coming through the walls. If this person is trying to connect to emergency services, the results could be tragic.

How first responders get coverage

After an initial investment by the federal government, billions of dollars are currently being spent by AT&T to build the new FirstNet network, bringing prioritized broadband telecommunications to the nation’s first responders. State and local governments are also investing to upgrade equipment. This new network is using a dedicated spectrum band (Band 14, also known as the Upper 700 MHz D-block) and also provides prioritized access to the AT&T commercial bands as needed during an emergency.

With fewer users compared to a commercial network, the FirstNet network will experience less congestion and, therefore, a higher signal quality than those serving hundreds of millions of users and devices.

With the addition of High-Power User Equipment (HPUE) Power Class 1, the FirstNet devices can transmit on Band 14 at up 31 dBm. This is a significant increase from the standard 23 dBm (Power Class 3). This can improve FirstNet coverage in fringe areas by up to 80%. Specifically, the ability for the cell site to better “hear” the user equipment can be the difference between a dropped or completed VoLTE call, delivered text message, or the transfer of mission critical data.

While FirstNet is being built into the robust system that has been promised, first responders still use their proprietary Land Mobile Radio (LMR) networks as their primary means of voice communication. Portable cell sites are also available in some circumstances to supplement wireless coverage where needed.

What’s being done to help the public

A significant federal effort has been underway during the past decade to improve wireless coverage in rural areas, but poor wireless coverage can be experienced in big cities as well. The wireless networks were originally designed to work well in a “mobile” environment – namely outdoors while in moving vehicles or walking. As indoor usage has grown, the networks have densified and greater efforts have been made to provide a signal strong enough to penetrate buildings.

Most single-family residential structures will typically be made from materials such as lumber and brick which the chart above shows as contributing to a minimal loss of signal. Buildings with a greater population density, such as multi-family residential and high-rise commercial structures, will typically employ thicker construction material in order to achieve the strength required to bear the weight of multiple floors.

Even where signal strength is strong, high demand on the network can impact user experience. These larger buildings mean more network users per square meter and that, in turn, creates added strain on signal quality. Wrap that building in eco-friendly low-e glass and poor wireless service shouldn’t be a surprise.
outside_signal_strength_philadelphia
The above image from Ookla’s Cell AnalyticsTM portal depicts a gradient heatmap of the outdoor signal strength provided by the Verizon Wireless network in downtown Philadelphia. The crowdsourced readings are averaged over the past twelve months. Red and orange represent a very high signal strength, whereas green to blue represent a lower signal strength. It is clear that Verizon has made significant investments in their Philadelphia network.

Providing high quality indoor coverage is much more difficult. Over the same twelve-month period, using Cell Analytics Pro building layers, we can view the same area in downtown Philadelphia with each building outlined in a color representing the average signal quality from readings captured inside each structure. It is clear that many buildings show an average signal quality rated as poor. Every mobile operator experiences these difficulties.
inside_signal_strength_philadelphia

How we can solve this public safety dilemma

Understanding which buildings fall short of providing adequate service can assist local governments in working with building owners and mobile operators to make needed improvements. This falls into two broad categories: First Responder Push to Talk systems and Commercial Mobile Services used by both Public Safety Agencies and the general public.

The solutions used today for First Responder Push to Talk systems are Distributed Antenna Systems (DAS) and signal boosters. For commercial mobile services, DAS, Booster Systems and Small Cells can be deployed based on individual use case. CBRS is a future Private LTE offering that is currently being developed and deployed in the United States.

Distributed Antenna Systems (DAS)

There have been solutions on the market for many years now, but the economic viability varies depending upon the use case. A DAS effectively deploys a miniature cellular network throughout a structure. DAS are very effective and have been deployed in large buildings, arenas and stadiums, but they are not appropriate for smaller buildings.

Signal boosters

Many companies make boosters that can capture outdoor signals from a nearby tower site then route them to repeaters inside of a building. This can solve a problem with signal strength and is more common for Public Safety LMR than cellular. This solution tends to be less expensive than installing a DAS network. However, if there is a need for higher capacity, a signal booster can actually exacerbate an issue by routing additional traffic to a cell site that may already be overloaded.

Small cells

Small cells are much in the news. Those being mounted to streetlights and other municipal structures are meant primarily to increase outdoor coverage at the ground level. This is particularly true with the new millimeter wave spectrum (extremely high frequencies) being used for some 5G deployments. These deployments will greatly improve coverage and quality on sidewalks and in vehicles, but mmWave is not designed to penetrate buildings.

Small cells can also be installed indoors, greatly improving floor by floor coverage in taller buildings. Using high-band (mmWave) spectrum also means that the high efficiency windows can block signals from escaping, lessening the chance that a small cell within one building would leak signal that could interfere with a different system in a neighboring building.

CBRS

The recently approved CBRS (Citizens Broadband Radio Service) technology promises to bring private LTE service to commercial buildings. Instead of depending upon the national wireless operators to provide a strong indoor coverage, an enterprise can deploy a solution to meet their specific needs, much like they have done with Wi-Fi.

So, what do most of the solutions above have in common? They are often deployed by the building owners, managers or commercial tenants. Although we will certainly see the mobile operators deploy solutions where the ROI justifies the cost, it will be up to the organizations that use wireless services every day in their businesses to underwrite the expense. The game-changer with CBRS is that a significant portion of the spectrum is unlicensed, therefore, coordination with a wireless operator is not required.

The game-changer with CBRS is that building owners have an opportunity to own and control the spectrum inside their own buildings, giving them more control over the quality delivered to their tenants and visitors. They will also have more visibility into call patterns and other data usually available only to the wireless carriers.

Building codes need to change

If the goal is to improve safety by ensuring the callers can reach 911 in an emergency and that first responders can maintain adequate coverage when being called to an emergency, then building codes must reflect this need. Sprinkler systems were initially installed to protect property from damage. The first fire code for sprinkler systems was written in 1896. As statistics began to show the death rate in buildings with these systems were dramatically lower, they became required in new construction. The requirement to retrofit existing buildings with sprinkler systems varied greatly from city to city and state to state.

As the cost to deploy indoor coverage technology declines, public safety officials within each local government should be considering how to implement code changes that will improve access to emergency communications. This process will take many years, so it is important to have empirical data to help prioritize which structures are most at risk. This may be a national issue, but it will be solved at the local level, one building at a time.

My thanks to John Foley at Safer Buildings Coalition for his assistance on this article.

Ookla retains ownership of this article including all of the intellectual property rights, data, content graphs and analysis. This article may not be quoted, reproduced, distributed or published for any commercial purpose without prior consent. Members of the press and others using the findings in this article for non-commercial purposes are welcome to publicly share and link to report information with attribution to Ookla.

| October 8, 2019

Investigating Internet Speeds on Tribal Lands in the U.S. and Canada

Ookla is headquartered in Seattle, traditional home of the Duwamish and Coast Salish people, where the mean download speed over fixed broadband is 142.67 Mbps. Thirty-six miles to the south on the Muckleshoot Reservation, it’s 101.85 Mbps. Meanwhile, the Makah Indian Reservation, located on the remote Washington coast, sees an average download speed of 5.91 Mbps.

Who gets fast internet speeds and who doesn’t involves a blend of factors (including geography, population density and economics) that is unique for every location, including individual Native American reservations and First Nations reserves. As Indigenous People’s Day nears, we are interested in how those factors combine to affect mobile and fixed broadband speeds on tribal lands across the U.S. and Canada. Because we are experts in internet speeds, not tribal policy, we are also offering our full tribal data set at the end of this article for anyone who would like to do further analyses.

Mobile speeds vary widely between tribal lands

Fastest-Mobile-Speeds-1

The Capilano Indian Reserve has the fastest mean download over mobile speed of all the reservations and reserves we examined. Also called X̱wemelch’stn, this community near Vancouver, B.C. is the most densely populated reserve of the Squamish Nation and its mean download speed on mobile is 27.8% faster than nearby Vancouver, B.C.

The Rumsey Indian Rancheria of the Yocha Dehe Wintun Nation virtually ties for the second fastest mean download speed over mobile. Located in California’s Coast Range, this reservation contains a casino and resort and experiences a 93.7% faster average than the state of California. With a nearly identical mean download speed, the Pala Reservation of the Pala Band of Mission Indians also houses a casino and resort and an average download speed 93.6% faster than California as a whole.

Except for Kahnawake, which has a 3.8% slower average mobile download speed than nearby Montreal, Quebec, all of the reservations and reserves on this fastest mobile list have download speeds at least 26% faster than the nearby non-tribal areas (adjacent large cities or encompassing states or provinces) we compared them with. Muckleshoot is 71.2% faster than Washington. Compared to California, Rincon is 49.9% faster, Morongo 37.1%, Twenty-Nine Palms 31.3% and Viejas 27.6%. The Mashantucket Pequot Reservation was 26.3% faster than Connecticut.

All ten of the reservations and reserves with the fastest mobile download speeds are home to large commercial enterprises including shopping centers, casinos, resorts and even a data center. Six are in California. Three are located in or adjacent to large cities.

Slowest-Mobile-Speeds-1

Our list of reservations with the slowest mobile download speeds contains some of the largest (by area) reservations in the U.S. A few have tribal enterprises including mining, casinos and tourism; most do not. All are located relatively far from large cities.

The Navajo Nation Reservation has the slowest mean download speeds over mobile of all the reservations and reserves we examined. This largest reservation in the U.S. shows an average mobile download speed that’s 80.8% slower than Arizona as a whole. Santa Clara Pueblo, located north of Santa Fe, New Mexico, has the second slowest mobile download speed on our list and is 68.5% slower than the state of New Mexico. White Earth Reservation is the third slowest on mobile, coming in 71.1% slower than the state of Minnesota.

The reservations and reserves with the slowest average mobile download speeds are at least 43.5% slower than neighboring non-tribal lands. The Allegany Reservation is 63.8% slower than New York. The Lummi and Colville Reservations are 63.3% and 63.0% slower, respectively, than Washington. In Idaho, the Nez Perce and Coeur d’Alene Reservations are 52.8% and 45.1% slower, respectively, than the state average. The Blackfeet Indian Reservation is 46.6% slower than Montana, and Pueblo Taos is 43.5% slower than New Mexico.

Fixed broadband speeds show even more variation

The ten reservations and reserves with the fastest fixed broadband download speeds are mostly in or directly adjacent to large, urban areas. All of the six reserves or reservations with the fastest fixed broadband download speed are in Canada with an additional seventh Canadian reserve in the top ten.

Fastest-Fixed-Broadband-Speeds-1

Cole Harbour 30, a Mi’kmaq reserve, has a very fast mean download speed over fixed broadband. It is located in the municipality of Halifax, Nova Scotia, a city we’ve previously found to have some of the fastest fixed broadband speeds in Canada. Cole Harbour beats Halifax’s speed by 213.1%.

Second-place Wendake is located within Quebec City, Quebec and has a fixed broadband mean download speed that’s 26.2% faster than the capital city. The Squamish Nation reserve of Seymour Creek 2 has the third fastest download speed among reserves and reservations. Located near North Vancouver, the reserve’s download speed is 24.0% faster than that of their neighboring city.

Most of these reservations and reserves show faster average download speeds over fixed broadband than comparable geographies — large cities when nearby and states or provinces when not. Cowichan 1 and Tsinstikeptum 10 are 19.7% and 16.8% faster than British Columbia, respectively. Burrard Inlet 3 is 11.2% faster than North Vancouver. The Hollywood Reservation is 5.5% faster than Hollywood, Florida, and New Songhees 1A is 1.7% faster than British Columbia. However, Salt River Reservation is 16.0% slower than neighboring Scottsdale, Arizona and the Puyallup Reservation is 4.5% slower than nearby Tacoma, Washington.

Slowest-Fixed-Broadband-Speeds-1

The slower list is evenly split between Canada and the U.S. and is mostly made up of rural and/or isolated reserves and reservations.

The Munsee-Delaware Nation No. 1 has the slowest mean download speed over fixed broadband of all the reserves and reservations we analyzed. Located in southwest Ontario, the reserve’s download speed is 99.4% slower than the provincial average. Close to Redwoods National Park in California, the Yurok Reservation is the second slowest for fixed broadband. It also has a mean download speed 96.9% slower than the state of California. Eskasoni 3 has the third slowest download speed over fixed broadband and was 95.6% slower than the surrounding province of Nova Scotia.

Fixed broadband download speeds on the rest of the tribal lands on this list are similarly slower than their comparable geographies: from the Chippewas of Georgina Island First Nation (96.1% slower than Toronto, Ontario) to the Cheyenne River Reservation (92.2% slower than South Dakota). Whitefish Bay 32A and Peguis 1B are 94.5% and 93.5% slower, respectively, than Ontario as a whole. In Washington, the Makah Reservation and the Quinault Reservation are 95.1% and 94.5% slower, respectively, than the state average. Annette Island Reserve is 93.6% slower than the state of Alaska.

There are so many reservations and reserves that we could not cover them all here. However, we are making our full CSV available for download if you’d like details on mobile and fixed broadband speeds on other tribal lands. This file includes June-August 2019 data for all locations that have 30 or more samples.

Ookla’s mission is to help make the internet better, faster and more accessible for everyone. We hope that by sharing this data we can contribute to conversation about internet equity for tribal lands. If you publish anything based on this data, please credit Speedtest IntelligenceTM as the source.

Editor’s note: The CSV was updated on December 10, 2019 to correct a sorting error that resulted in the misattribution of states.

Ookla retains ownership of this article including all of the intellectual property rights, data, content graphs and analysis. This article may not be quoted, reproduced, distributed or published for any commercial purpose without prior consent. Members of the press and others using the findings in this article for non-commercial purposes are welcome to publicly share and link to report information with attribution to Ookla.

| October 31, 2017

LinkNYC Proves Public Wi-Fi Can be Free, Fast and Far-reaching

Two years ago, LinkNYC set out to provide New Yorkers with free Wi-Fi using a network of Links that offer “superfast, free public Wi-Fi, phone calls, device charging and a tablet for access to city services, maps and direction.” The experiment worked, and we’re here to report on how fast the network really is, whether it holds up to heavy use and which parts of the city are seeing the most benefit.

What is LinkNYC?

LinkNYC uses kiosks (“Links”) to provide internet connectivity to the five boroughs (Manhattan, Brooklyn, Queens, the Bronx, and Staten Island). The project resulted from a competition initiated by the New York City government to find a new use for the City’s payphones. Service at the first 500 Links was available by the end of 2015. One thousand, one hundred and sixty-four are active as of this writing with another 215 installed and awaiting final connection. The project aims to have 7,500 Links online by the end of 2023. Each link is connected to the LinkNYC fiber network and is capable of speeds up to 1 Gbps. The Links use 802.11ac Wave 2 access points with support for 4×4:4 MU-MIMO.

LinkNYC was created by CityBridge, a consortium of companies including Intersection (a portfolio company of Alphabet Inc.’s Sidewalk Labs), Qualcomm and CIVIQ Smartscapes. Beyond a $300 million capital investment from CityBridge, LinkNYC is funded through advertising revenues from ads shown on screens on the Links themselves and is projected to create more than $500 million in revenue for the City of New York.

How fast is fast?

Speedtest data reveals LinkNYC offered up a mean download speed of 158.98 Mbps and a mean upload of 123.01 Mbps in September 2017. This free network is faster by 74% and 165%, respectively, than New York City’s average download and upload speeds over mobile Wi-Fi. New York’s mobile Wi-Fi download and upload speeds for all networks combined are respectively 249% and 312% faster than those over cellular networks. That means LinkNYC users are seeing download speeds 511% faster than they would be if they were connecting on cellular networks. LinkNYC’s mean upload speed was 996% faster than cellular.

New York City Speedtest Results
September 2017
Connection Type Mean Download (Mbps) Mean Upload (Mbps)
LinkNYC Mobile Wi-Fi 158.98 123.01
Mobile Wi-Fi 90.89 46.27
Cellular 25.98 11.22

Even better, we can see that the speeds available on LinkNYC are improving rapidly. Though the Speedtest results fluctuate somewhat on a day-by-day basis, on January 1, 2017, the average download speed was 95.70 Mbps and the upload was 81.50 Mbps. On September 30, the daily average download speed peaked at 201.00 Mbps and the upload speed was 177.10 Mbps.

Sample Speeds by Day

These daily speeds likely reflect limitations of mobile devices, not those of the LinkNYC network.

One thing to note is that while the LinkNYC network is reportedly capable of 1 Gbps speeds, most mobile devices are not. The Speedtest results graphed above are likely limited by a world where the peak achievable speed for an iPhone 5 maxes out at around 100 Mbps and an iPhone 7 at around 560 Mbps.

To top it off, LinkNYC’s daily speeds are faster still than the averages for the U.S. as a whole. In September, we saw a mean download speed over mobile Wi-Fi of 66.25 Mbps in the U.S., while cellular came in at 23.98 Mbps. Mean upload speed over mobile Wi-Fi for the U.S. was 23.59 Mbps, while cellular was 8.38 Mbps.

Is it secure?

LinkNYC actually has two networks — one called “LinkNYC Free Wi-Fi” that’s open and available to all and “LinkNYC Private.” Hotspot 2.0-compatible devices are automatically prompted to join this second network which is protected using WPA2 encryption and Hotspot 2.0 technology which facilitates cellular-like roaming on capable devices. That means that once a phone joins the “LinkNYC Private” network, the phone’s connection to the network is handed off from access point to access point as a user moves throughout the city. Most carriers have offered Hotspot 2.0-capable phones since 2014 (T-Mobile has for much longer).

We can’t say for certain that the private network is secure, but the fact that it is encrypted likely makes it more secure than your average coffee shop connection. In light of the recently discovered global WPA2 vulnerability, we asked LinkNYC how vulnerable their network was. The response was: “The Link private network is not at risk. The combination of access point firmware and the network configuration assure the integrity of encrypted WPA2.”

Speeds over the two networks are similar with downloads over “LinkNYC Private” Hotspot 2.0 network just 9.4% faster than those on “LinkNYC Free Wi-Fi”. This could be due to the fact that newer phones are both more likely to offer faster speeds and to be Hotspot 2.0-ready as well as the fact that fewer users are accessing this second network.

LinkNYC Speedtest Results
January 1 – October 11, 2017
Connection Type Mean Download (Mbps) Mean Upload (Mbps)
LinkNYC Free Wi-Fi 131.84 100.97
LinkNYC Private 144.21 106.98

Do speeds hold up during peak periods?

Yes, yes they do. The graph below shows that even during periods with the highest volume of tests (from 3-4 p.m.), the LinkNYC network maintains peak download speeds. The slowest download speeds, seen from 5-6 a.m. and 8-9 a.m., still maintain an average download speed over 120 Mbps.

NUMBER OF TESTS BY DAY GRAPH

The far-reaching impact of LinkNYC

How many people are using it?

According to LinkNYC, the network sees 300,000 users per day. We saw 19,289 Speedtest tests taken with Speedtest over the LinkNYC network from 7,594 unique users from January 1, 2017 through October 11, 2017. This compares with the 2,436,040 Speedtest results from 322,851 users on all mobile Wi-Fi networks in New York City during the same period.

Where can you get it?

As described above, Links are rolling out rapidly. You can see the current map of available Links here, but we found that the map at today’s writing corresponds closely with the map of Speedtest results we’ve seen over the past year.

SPEEDTEST RESULTS ON AVAILABLE LINKS

Much of Manhattan is covered by Links, with the exception of Washington Heights north of the George Washington Bridge and lower Manhattan below Worth St. Links are sparser in the Bronx with nothing north of West Kingsbridge Road or east of Webster Ave. (north of 161st) and Third Ave. (south of 161st).

The alignment of Links along major thoroughfares becomes more obvious in Queens and Brooklyn. In this area, we see a large gap around Bushwick and Ridgewood with nothing south of 39th St. and Clarendon Rd. A whole swath of northeast Queens (bordered by Grand Central Parkway) also exists in a Link-free zone.

Meanwhile, in Staten Island you’ll find Links on sections of Hylan Blvd. and Richmond Rd. with a few scattered in between.

Links are required to have a range of at least 150 feet, although a spokesperson for LinkNYC reported a Speedtest result as fast as 80 Mbps at 400 feet. At either distance, people living in neighborhoods with a more sparse Link distribution may not have access near home.

PERFORMANCE MAP

The map above shows that speeds vary not by neighborhood but by distance from a Link.

Can I use it at home?

If you’re lucky enough to live near a Link, you actually can use it for your home internet; there’s nothing in the terms of service to say otherwise. But, unless you’re sitting within the unassisted range of a Link, we wouldn’t recommend it.

We were able to identify Speedtest results that were using extenders and/or repeaters to rebroadcast the service farther into a building, and those tests showed speeds 57% slower than New York’s 90.89 Mbps average for downloads over mobile Wi-Fi.

LinkNYC Speedtest Results
January 1–October 11, 2017
Connection Type Mean Download (Mbps) Mean Upload (Mbps)
Unsampled Results from Extenders and Repeaters 38.87 30.71
Overall LinkNYC Results 158.98 123.01

Approaches to municipal Wi-Fi

LinkNYC’s ad-based, public-private partnership is only one approach to municipal Wi-Fi. In Leiden, Netherlands, a non-profit organization set up an independent, community-based network run by volunteers. Venice offers free Wi-Fi to all residents and Longmont, Colorado offers free municipal Wi-Fi in several locations, mostly parks, as part of their Gig City initiative.

LinkNYC isn’t the first free Wi-Fi network, but it’s one of the largest. And, despite some bumps along the way, it’s successfully bringing fast, free public Wi-Fi to the masses. The project is so successful that CityBridge has just extended the same model to London as InLinkUK.

Are you enjoying free mobile Wi-Fi on either LinkNYC or InLinkUK? Take a Speedtest on Android or iOS so we can share details about the performance of these networks over the long haul. Select the CityBridge server when taking your test to get the most accurate results.

Ookla retains ownership of this article including all of the intellectual property rights, data, content graphs and analysis. This article may not be quoted, reproduced, distributed or published for any commercial purpose without prior consent. Members of the press and others using the findings in this article for non-commercial purposes are welcome to publicly share and link to report information with attribution to Ookla.