My Rogers Toronto area LTE test

I had the chance to test out Rogers Toronto area LTE network last Friday. I've been bugging Rogers for a while to let me test their Toronto LTE network before it officially launches. I finally convinced them to let me try it with a few conditions... a) it took place in one of their red Rogers vans and b) I couldn't take the service with me after the drive.

Anyways, there are many problems with my test which actually caused me to ponder whether I should write anything at all. First off, Rogers could have taken me on a route that is flattering to Rogers. I only got to run the test for about 45 mins. Their LTE network is not public yet so I basically have the entire network to myself but at the same time it may not be ready yet. There could be some last minute network tuning that needs to be done. The test was run around 1pm when networks are probably not that loaded since everyone's at work.

Before I tell you how I tested, let me explain what advantages an end user should notice from LTE over HSPA. You'll notice faster downloads, uploads and lower latency. Of course there are many many variables which can make this not true. For example if their LTE network is overloaded it could result in worse results when compared to HSPA. It's not likely but it could happen.

LTE's sexiest feature is that it boasts download speeds of up to 150mbps. In the case of Rogers, I believe their implementation currently maxes out at 75mbps which is still very fast for a wireless connection. Since I only had 45 mins, I decided to test the maximum download. I decided to use Firefox to download multiple ISO's simultaneously from multiple sources in an effort to max out the line. Next I used a program to log the network throughput every second.

The approximate path we took:

Started at 333 Bloor St East and headed east
Left onto Mt Pleasant
Left onto St Clair ave
Right onto Yonge St
U turned when we realized there was construction on Yonge and went back to St Clair
Left onto Avenue Rd
South on Avenue to Queen's Park
Left onto Wellesley St
Left onto Jarvis back and back to 333 Bloor St.

Here are the results:


I observed a maximum speed of 60mbps, a minimum speed of 0mbps. I also listed what percentage of the results were less than 5mbps, 4mpbs, 3mbps, 2mbps and 1mbps to help give you a feel of how bad the bad results were.

I also brought a Bell LTE stick along with me for the ride but for reasons I'm still looking into, it wouldn't connect to Bell's LTE network <insert conspiracy theory here>.

Here are my observations. It goes without saying but on an empty network, LTE is fast. It's so fast it kept finishing my downloads so there were actually a few periods when there was no network activity (I removed them so they don't show up on the chart). While LTE is really fast, the Bell LTE stick which was operating in DC-HSPA+ was also really fast. I'm still looking through the data but the Bell stick hit 30mbps+ multiple times. So while LTE is fast so is DC-HSPA+. It makes me wonder what real world applications actually require LTE. An analogy could be a road car with 800hp vs a car with 600hp. Both have enough power.

I was also thinking of the cost to use these. After my test I took the Bell LTE stick (which was still in DC-HSPA+ mode) and downloaded around 10GB in just over 2 hrs. It's a completely unrealistic test as current data plans don't really allow for this sort of use. To continue the car analogy, it's akin to having a car with 600hp but only 1 liter for fuel in the tank.

In the end I'll have to wait until Roger's LTE network actually launches before I can do a meaningful comparison. Then I'll be able to drive around and run these tests in my own car on a commercially available network.