The Way of Wireless

Wireless is great, it allows me to connect my printer, PC tablet, set top box and of course use my mobile for voice and data. But if I need a proper connection, I will still connect to wired connection whenever possible. The trouble with Wi-Fi in my neighborhood is everybody else also has Wi-Fi and the problem with that is the limited number of wireless channels available.

The end result is that despite my best attempts of moving my access points around and channel hopping, inevitably my neighbours will configure their access point to use the same channel, and that causes my Wi-Fi to choke. The problem is made worse with wider bandwidth 802.11n or the 105Mbps access points, better antennas, range boosters, everyone is fighting for the same limited set of channels in the same chunk of noisy spectrum, which is also shared with my microwave oven. In the past, a single 802.11g access point running at 54Mbps would happily service all the wireless devices in my home.

Today, I have two access points just to cut through the noise. So, for myself at least wireless is used for wireless devices that don't have a wired connection. Where I live the case with 3G is only slightly better, and depends on where I am. My service provider has great coverage in some areas and terrible service in others, as a result mobile broadband will jump between UMTS (3G) and EDGE (2.5G) causing whatever download or upload in progress to stop completely in some cases.

Throughout the Middle East and Africa, I have found the situation is similar. Usually these networks are a hybrid of both legacy and new mobile switching centres, connected with microwave links, copper and fibre technologies. Indeed in other countries I have visited such as Istanbul, Turkey, where the deployment of 3G UMTS data services is fairly recent, the 3G Internet speed is so fast with some operators that you can forget you are connected with a modem dongle.

The reason behind this is twofold, an all fibre Ethernet backbone for backhauling radio traffic, and relatively expensive data plans, which cause users to download only when it is really necessary. In other big cities, where the backhaul infrastructure is already upgraded, the mobile wireless situation is slightly different, simply too many people simultaneously attempting to access services using the same limited radio spectrum. In these areas LTE (4G) has promised to solve these issue by using more efficient radio modulation techniques.

However, I can't help but think that in these large markets the current roll out strategy of LTE will be akin to shifting piles of sand. Initially, the early adopters will enjoy the speed and improvements LTE brings, but soon the masses using the congested 3G network follow and coupled with the explosive demand for wireless, we will need a new approach to solving the growing bandwidth demands.

The approach which I believe will be the ultimate saviour of the wireless provider, will be low cost, low power pico and femto cells. In next generation mobile networks, it will be critical to get the wireless traffic into the ground and back into a fibre network as soon as possible, thus reducing the air transmission load on cell towers.

In addition, I expect to also see per subscriber traffic prioritisation to start becoming popular. If you look at the massive investment mobile operators are putting into their network today, you will see that although the speed is increasing significantly, the ARPU is either static or only expected to increase slightly. In the future, we will see best effort data services, offered alongside both guaranteed bandwidth and guaranteed latency plans, all of which is made possible with all IP LTE networks.

To make any of this possible however, frequency spectrum must be harmonised to ensure seamless handover between cell technologies. The amount of available bandwidth is dependent on the width of the available frequency spectrum. In many countries, pockets of frequency spectrum is still allocated to different government agencies, legacy pager networks and analog television.

Operating a 4G network around these obstacles is inefficient and in worse case scenarios, only marginal improvements over existing technologies can be seen. All these thing must occur together to meet the demand of the growing wireless broadband community, since failure to do so could end up with wireless mobile broadband being no better than my Wi-Fi connection today.