Sprint Nextel's Network Vision Neglect

Mistakes made in upgrading a network to LTE

I like Sprint.

I like Sprint's pricing and corporate policies. I like their strong support of MVNO's like Ting. I like their partnership / support of Google projects, such as native Google Voice and not blocking Google Wallet. I like their credit union member discounts. I like the fact that they care enough about their service to offer in-market roaming (something T-Mobile won't allow).

I like what Sprint stands for. Sprint is the largest wireless internet service provider without a monopoly to fall back on. They are the largest company left preventing AT&T and Verizon from owning all wireless internet access in the US. Sprint doesn't have a wireline phone service to fall back on (like AT&T and Verizon). Sprint doesn't have a internet/television offering to fall back on (like AT&T has with U-Verse, or Verizon has with FIOS and Comcast)

I like Sprint. Which is why, it is with heavy heart that I write this article. While there's a large number of things to like about Sprint, there's a major problem.

Sprint simply doesn't maintain a solid network in Grand Rapids. They chronically make poor decisions that actively hurt the service they provide. When a problem occurs, Sprint is usually faced with two decisions; one to fix the problem long-term, and one to band-aid the problem with a temporary, unsustainable patch. Almost universally, Sprint opts for the "band-aid" solution.

I write this article, not because I want to rag on Sprint. This is not as an attack on a weak player. It's an intervention. There are major problems here that need to be addressed, problems that have been ongoing for years and are not yet resolved.

I'm confident Sprint could fix these issues, if they chose. There are markets where Sprint service works. But Grand Rapids is not one of those markets, and until it is, these points stand.

What's the problem?

Sprint faces two major problems right now. The first revolves coverage -- they simply don't cover the areas they claim to cover, and are expending no effort to improve this. The second revolves around data service - they simply don't offer the data service they claim to provide.

Sprint's service is so bad in Grand Rapids, it's been the subject of multiple public inquiries in various local media for years.

WOODTV investigates Sprint service, March 2011

The coverage issues stems from a lack of cellular sites. Sprint doesn't operate enough cell sites per square mile in the Grand Rapids metro area to provide usable service. The data service issues stem from a lack of cellular sites, and a lack of backhaul. (Sprint doesn't purchase enough internet access at their cell sites to support the number of subscribers they've sold internet access to)

LTE just lanched in Grand Rapids. LTE will fix things!

Sprint's LTE service "launched" this past week, in Grand Rapids, Michigan. Launched is the term they used in the Sprint LTE Press Release, but it doesn't mean what you would think. When Sprint claims a market is "launched", it only means that they claim LTE coverage in 50% of the outdoor areas of that market.

I've been examining their LTE network performance using a Motorola Photon Q LTE, and the results fit in roughly with what most would expect. In rural areas, you see speeds between 6-8mbps with peaks around 20mbps (exactly as advertised). In the suburbs (Grandville, Wyoming, Walker, Kentwood, ect) speeds are between 2-4mbps with peaks near 8mbps. And in urban Grand Rapids, speeds average between 0-2mbps.

Typical Sprint LTE service in August 2013 at downtown Grand Rapids, Michigan

If your seeing those numbers and thinking, 'that seems really slow for LTE', you are not alone. Those speeds are identical to T-Mobile's "3G" HSPA+ 21 (which uses an identical 10mhz of spectrum in Grand Rapids as Sprint's current LTE network)

Effectively, Sprint's brand new LTE network is over capacity in Grand Rapids before it's even launched. Additional sites will come online in the coming months (Sprint's committed to upgrading every EVDO cell site to LTE), but even if they were all operational, speeds aren't likely to improve for urban users.

What about 'Network Vision', won't that fix things?

"Network Vision" is the term Sprint uses for it's current network modernization strategy. The hope is that by pushing through 4 years worth of skipped network upgrades at once, they can make up for lost ground between Sprint and the other 3 major carriers.

Network Vision contains a lot of good, useful changes. For example, Network Vision deploys LTE service on every Sprint PCS cell site, with additional backhaul to support 5x5 LTE. It also modernizes backhaul for the EVDO service, and adds support for 1x service on the 800 band (which should provide for improved voice and sms coverage).

Unfortunately, while Network Vision contains a lot of good changes, it hurts the overall network as much as it helps. Of the two major problems mentioned above, Network Vision fixes one (backhaul), but hurts the other (coverage). This is because, the Network Vision plan removes 30-50% of all the towers and coverage across all of Sprint's holdings in Grand Rapids.

Network Vision cuts 30-50% of towers?

Yes, it's true. For many years now, Sprint's been paying for additional cell sites and coverage in the form of Nextel. Nextel's network is 30-50% more dense (in terms of cell sites per square mile) than Sprint's network is. Additionally, Nextel's network used a lower band spectrum (800 vs 1900), which allowed for better signal propagation.

With Network Vision, almost all of Nextel's sites are being permanently closed down, with no replacement planned

Now, some of these towers were truely redundant (they were co-located on the same sites that Sprint PCS is on). But the vast majority of them arent, they were unique to the Nextel network, and filled in 'dead spots' that plague Sprint in Grand Rapids.

Below are a collection of maps comparing Sprint and Nextel cell sites. Notice how in every area, Nextel's sites were closer spaced, and often filled in well-known dead spots.

Comparing Sprint (yellow) and Nextel (red) sites as of Summer 2013 (click image for full size)
Comparing Sprint (yellow) and Nextel (red) sites as of Summer 2013 (click image for full size)
Comparing Sprint (yellow) and Nextel (red) sites as of Summer 2013 (click image for full size)
Comparing Sprint (yellow) and Nextel (red) sites as of Summer 2013 (click image for full size)

Keep those maps in mind when you read articles mentioning Nextel subscriber losses. It's easy to paint these folks as old, anti-upgrading individuals, when in fact, many of them depended on Nextel for service, and simply couldn't switch to Sprint, as Sprint wouldn't offer a remotely comparable level of service.

What needs to be done to fix Sprint's network?

Sprint's already making some good moves. Network Vision's improved radios, additional backhaul and rollout of PCS-LTE will improve service on those towers to at least 3G levels of it's competitors in the short term. It's plan to offer 800 LTE, and EBS LTE service should allow Sprint to roll out their first network truely deserving of the "4G" label.

However, in order for those plans to work, Sprint must add more cell sites. Their sites are spread out so far apart that PCS spectrum network has large neighborhood-wide coverage gaps. Adding EBS into that mix is a recipe for disaster.

Normally, I'd recommend they just use the Nextel sites. (They are already leased and operational. With additional backhaul and new radios, they could be perfectly serviceable). However, many of these sites have already been lost after the shutdown of the Nextel network. Leases have been left to expire without renewal. If Sprint can get these sites back, it would of great help. But it might be too late.

The last remaining hope is with Clearwire. Sprint is once again in a position where it's aquiring a network with a denser, stronger site layout than it's own. Sprint's once again at a choice where they can keep the increased density, or remove it.

I strongly urge to Sprint to keep the Clearwire network of cell sites. Every single Clearwire site should be upgraded to a full network vision site, running a full complement of 800, 1900, and 2600 services.

Additionally, I urge Sprint to invest in distributed antenna systems. AT&T, MetroPCS, and T-Mobile already use these in various areas to great benefit. Sprint would be wise to follow in their footsteps

Sprint's network is not a lost cause. These issues can all be fixed. But only if Sprint makes a major investment in additional cell sites.