Not only has internet on-the-go arrived, it’s already being taken for granted. According to research, the total number of mobile internet users worldwide is expected to grow at an annual rate of 16.6 per cent by 2015. This recent surge has mainly been fueled by products with better user interfaces (such as tablets and smartphones), the right kind of application ecosystem (where developers are free to be creative), solid and fool-proof underlying architecture, and of course, better wireless network connections. The current model for wireless communication relies heavily on two access methods:
- Using one’s mobile device to access the carrier’s data plan for internet.
- Using one’s mobile device to access Wi-Fi for internet.
What’s your plan?
If you are a cellular service provider, deploying cell towers is a lot more expensive than say, deploying Wi-Fi hotspots. Even if you deploy cellular towers with higher networking capacity (LIE, HSPA etc), you are still essentially building a thicker, dumb data pipe. Cellular providers are consistently being pushed to provide faster services. However, incentives like revenue per active subscriber (driven by the increasing demand for data), do not justify building a big agnostic network that costs each mobile operator in United States an average of 20 billion US dollars. Operators can’t even transfer these costs over to the consumers, because as we well know, consumers do not want to pay more for mobile internet.
The intelligent solution
The solution that mobile operators are increasingly looking towards lies in intelligent networks Operators also want the ability to priorities some of the traffic that goes through their network, like video over HTTP. This gives rise to second-order intelligence of traffic class identification. Perhaps operators want to harness such information to possibly charge more for accessing a video than accessing a simple webpage. So increasingly, we see such differentiation built into the network that would allow subscribers to pay less for data connectivity and operators to develop new revenue streams in their network. Another use case for traffic class identification is compiling network usage pattern data. This can be used for generating ad revenues by collaborating with application space to not only contextualize experience for wireless data users, but also enable targeted marketing. It is essentially what Google does by providing free services to serve ads, but in this case, it is on the network layer instead of the application layer.
Internet in a suitcase
What if the access infrastructure is not present at all? Recently, an open-source software called Serval emerged. The Serval project provides temporary and permanent mobile network solutions, and removes the need of cellular towers. It forms an ad-hoc network and allows users participating in the ad-hoc network to make calls, send text messages, and access the internet (subject to the availability of a gateway) using Wi-Fi. Though its development was done on the Huawei Ideas, other Android phones are also supported. Other successful demonstrations of what is called “Internet in a Suitcase” have also been conducted. The idea is to be able to pack all the required equipment to create a wireless hotspot in a remote location away from infrastructure networks, in one suitcase. This suitcase can then be used to provide access to users in a particular location, and use any outgoing internet connection, including satellite up-link. Both of these avenues describe an increasing trend towards openness. The idea is for the internet to be open and free for all. Service providers should be reduced to doing what they do best — providing dumb pipes. However, these goals are directly in conflict with more established methods. A cellular company does not only provide the network, but also markets the service and builds a business out of it. Unfortunately, most of the latest ‘ developments in mesh networks haven’t been properly developed into business cases yet; this limits their spread and usage. So a marriage of both worlds is needed for the next generation of access technologies.
Free internet in the making
Some glimpses of that union are visible in the rise of Wi-Fi offload. Wi-Fi offload is another form of network intelligence, where a cellular service provider can hand off your data requirements onto a local Wi-Fi hotspot. This enables a user to access one or more network(s) simultaneously — quenching the data demand. Next generation standards like 3GPP Release 11 show how this can be achieved, enabling operators to optimize the revenues on their networks by reducing the cost to serve more users. This, of course, is not the ultimate solution for a totally free-of cost internet. It is, however, a step in that direction.