I have to scratch my head when I read about people who rave about the 4G data speeds they get. First, I travel all over the country and I have yet to see a 4G data speed above 20 Mbps. And yet I’ve seen claims in various online forums for speeds as high as 60 Mbps. I’ve been in a number of major cities in the last six months and have not once seen speeds that I would consider fast.
Second, a report just came out from OpenSignal, a company that provides an app that maps cellular coverage. They collected data recently from 325,000 users around the world and used that data to compare the 4G networks in 140 different countries. Their data showed that the US has the 14th slowest 4G of all these countries at an average paltry speed of 10 Mbps.
Hungary, Denmark, South Korea, Romania, Singapore, and New Zealand have the fastest 4G, all with average speeds of above 25 Mbps, with New Zealand seeing an average speed of 36 Mbps download.
I often run speed tests, but the real way to test 4G speeds is by trying to open web pages I often use at home. I know it’s generally far more efficient to use an app rather than the mobile web, but I open web pages just to see how fast coverage is. It’s well known that speed test results can be rigged by your carrier who knows you are using a speed test site. What I generally find is that web pages that leap onto my screen at home seem to take forever to load on my cellphone, and sometimes they never load.
Why does this matter? I think it matters because there are tons of industry pundits who opine that our broadband future is wireless and that we don’t need to be investing in fiber. They say that wireless is going to get so fast that nobody will feel the need for a landline based internet connection. For a whole long list of reasons I think that argument is totally bosh. Consider the following:
- Cellular data speeds drop quickly with distance from the cell tower. Today cell phone towers are not situated for data coverage and were built to handle voice traffic. A cell tower can complete a voice call at a much greater distance from the tower than it can make a strong data connection.
- We could always build more towers to bring transmitters closer to people. But for those new towers to work they are going to have to be fiber fed, something that very few companies are willing to invest in.
- Cell phone signals don’t penetrate structures very well. I recently visited my dentist. In the parking lot I was easily able to read news articles on Flipboard. I then walked into the waiting room, which has big windows to the outside world, but the ability to read articles slowed down a lot. Then when I was taken back to an interior room that was only one room further from the outside, I couldn’t even get the app to open. This is not an unusual experience and I see it often.
- Cell sites can only handle a limited number of customers and they get overwhelmed and degrade if they get more demand than optimum. And the more bandwidth that is delivered, the easier it is for a cell site to reach capacity.
- The various swaths of spectrum used for cellular data each have their own unique limitations. In many cases the spectrum is carved into somewhat small channels (which was done before we conceived using the spectrum for data) and it’s very hard to cobble together a large wireless data path. It generally means linking several frequencies to a given customer data path, which is both complicated and somewhat taxing on a cellphone.
- Data caps, data caps, data caps. Let’s face it, as long as the cellphone companies want to charge $10 per downloaded gigabit then they cannot be a serious contender for anybody’s real life data usage. I estimate that my household downloads at least 500 gigabits per month at home and I don’t think we are unusual. If I was paying cellphone data rates that would cost me an astounding $5,000 per month. Even should they cut their rates by 90% this would still cost an eye-popping $500 per month. As long as cellphone data rates are 100 times higher than landline rates they are something you use to casually browse the news, not as a real internet connection.