FWA Cellular Speeds

One of the most interesting things about getting access to a lot of speed tests is that it provides a way to test broadband issues you always suspected but couldn’t prove. If you can collect enough speed tests, you might find proof of a lot of different things. For example, speed tests might show that a broadband network is slower in the evening than during the night – something that customers have always complained about. Speed tests might show that an ISP delivers speeds that are far slower than what an ISP claims on the FCC broadband maps.

I’ve been trying to understand the speed characteristics of FWA cellular wireless. I’ve been interviewing folks for a few years who have FWA wireless, and they all told me that speeds are fast for those living close to a tower but slower as the distance to the tower increases. For example, the first customer I talked to who was using the FWA broadband from T-Mobile is a farmer who had a T-Mobile tower on his property and got almost 300 Mbps download speeds. He was thrilled with the product compared to the much slower WISP he had been using. But when he recommended the FWA wireless to his neighbors, they received a far different bandwidth product. A neighboring farm a little over a mile away was getting speeds closer to 100 Mbps, which they also thought was good. But some farms further away said that the FWA broadband was too slow.

I heard similar stories from elsewhere, but it’s hard to make any universal statements about the FWA product based on a handful of anecdotes from different parts of the country. I recently got access to enough speed tests to understand the performance of the FWA cellular wireless product.

The map below shows a lot of speed tests from Verizon tower in a suburban county. The yellow dots on the map are the locations of actual speed tests. The colored circles on the map show the distance from a cell tower – with purple showing locations within a mile of the tower, red showing locations between 1 and 2 miles, blue/greed showing speed tests within 2 and 3 miles, and the surrounding white areas at more than 3 miles. I didn’t cherry-pick this particular tower as the best example – there are more than a dozen other Verizon towers in the same county that show similar speed test results. I must note that speed tests are not a prefect indicator of broadband performance, and there might be explanations behind some of the slower readings. But I have to think that seeing this same speed pattern around multiple tower sites is a good indication that this is how the technology works.

This map demonstrates what the farmer told me to a tee. There are some locations close to the tower getting 300 Mbps. Customers just over a mile from the tower are getting slower speeds, with the highlighted ones around 75 Mbps. By the third mile band, speeds have dropped a lot closer to 25 Mbps download, and outside the three-mile circle, speeds drop significantly. There is no easy way to tell if the customers with slower speeds are buying FWA wireless, which uses the spectrum that Verizon labels as 5G, or the older Verizon hotspots that use traditional LTE spectrum.

On the FCC map in this county, Verizon reports two speeds – 300 Mbps or 50 Mbps. It’s not easy to understand how Verizon makes the distinction, but it seems like locations for a fairly good distance around towers are claimed at 300 Mbps.

Somebody who doesn’t understand the FCC mapping rules might think that Verizon is breaking the rules by reporting 300 Mbps speeds in places where actual speeds are a lot lower. But the FCC allows ISPs to report marketing speeds for the FCC maps as long as Verizon is advertising the claimed speeds. But that doesn’t mean that the Verizon FCC reporting is ethical. Customers who might refer to the FCC map when looking for an ISP, or customers that see Verizon advertising are hoping to get something close to the 300 Mbps speed – and many will not.

I have some major concerns about cellular FWA technology related to the upcoming BEAD grants. First, any state broadband grant offices that accept the claimed Verizon speeds in the FCC mapping might not award any grants where a fast FWA speed is claimed. That would be a travesty if folks who can’t get speeds of at least 100/20 Mbps with FWA are denied another broadband option.

It’s also possible that the cellular companies will challenge grants that come close to their towers. I knew this was likely going to become an issue the day that the NTIA said that it considers wireless broadband using licensed spectrum to be broadband for purposes of the BEAD program.

It’s also possible that Verizon, T-Mobile, AT&T, and others will try to win BEAD grant funding using this technology. At least in this county, there are very few customers outside of one or two miles from a tower who can get the 100/20 Mbps required for BEAD grants.

I hope that state broadband offices take a hard look at this. Many of them have purchased detailed speed test data, and they can search around towers in the same manner done above. I don’t think it will take much investigation for them to be convinced that FWA cellular broadband can meet the speeds required for BEAD – but only for short distances from cell towers. Broadband offices should also take note that both Verizon and T-Mobile warn customers that speeds can be throttled any time there is increased demand for bandwidth from cellphones.

I am not busting on the cellular FWA technology. If I was in a rural area without a good broadband alternative, I’d buy this product in a second. But I’d be unhappy if I was hoping for 300 Mbps and got 25 Mbps. What is being deployed today is the first generation of the technology, and I assume that it will improve over time. My only concern is the timing of the rollout of this new technology and how it might negatively affect an already complicated BEAD grant process.