How Much Throughput is Enough?
With all the talk of 3G/4G/5G wireless networks and the speeds they provide, the question really comes down to how much speed is enough? With mobile network operators (MNOs) data caps limiting overall data utilized, faster speeds really just mean reaching the limit faster.
There is something to be said for getting information quickly, but at some point networks seemed to have crossed the point of what we truly need as subscribers and what their wireless network are now actually providing. With 5G promising 1Gbps speeds, they seemed to have moved past the consumer phone business and more into the 4k video stream business.
There also is a marketing hurdle to cross that the speed delivered to your phone is closing in on the speed delivered to your house. This is advertising. The reality is a phone is a single device, where the home is a small network of interconnected devices each requiring a piece of the bandwidth provided (TVs, computers, tablets, smart devices, radios, etc). Each interconnected device may not require the entire 40MB pipe for example, but combined on a Friday night with Netflix, the DVR recording two shows, and two kids surfing the internet listening to Pandora upstairs do.
Wireless Network Operators spend billions each year expanding their network, deploying next generation technologies, and adding network capacity overall. With prices for consumers falling, network operators will no longer be able to maintain the open max pipe. This doesn’t imply throttling, it just means they will no longer push for the max speed for each user, rather ensure an acceptable speed is available to each user. For example the typical LTE speed tests today are often well over 25 Mbps. If the mobile user really requires 2.5Mbps for 99% of their activities, why expand the network to provide more?
Figure 1: Typical Speed Required for Internet Activities
Often the answer given is the future. Is the future for MNOs truly higher bandwidth applications or is just exponentially more connected devices tying up their bandwidth. Many of the future applications are monitoring devices, vehicles, and smartphone applications which actually require constant connection, but relatively small throughput requirements. It is this future of devices that actually make us look at the quality of the data network from a user point of view.
Data traffic does not have the same instantaneous perception by the customer compared to the visibility as voice traffic. Consumers all hear poor audio, dropped calls, and having to redial after a call failure of voice. To the consumer, there is no “data drop”, there is not latency issue, there is only whether the device did what was needed to be done in a reasonable time. Dimension of the mobile networks needs to not look at the maximum speed available to all consumers, rather than the minimum acceptable level is being met for current users during peak times. During non-peak times, subscribers may be getting 20Mbps+ as they are the only users however, at peak times there will still be enough that they don’t see problems at around 5Mbps.
The question really comes on whether the MNOs would truly like to replace the existing terrestrial services such as cable, fiber, and DSL. For modernized areas, this may not be financially viable option for MNOs, however in current developing nations and rural areas, this may truly be the solution. Some bandwidth is better than no bandwidth. With the internet providing a greater value in terms of education and opportunity for these as well as all other areas of the world, faster speeds may be required, but the question would still remain how much bandwidth would be required? A service cell tower may need to gigabit back-haul, but each individual end user still will not require all of it.
As there are more connected devices attached to the wireless network requiring bandwidth and spectrum becoming less available, operators will need to control capital costs. The “Fastest Network” claim will have to move to a different marketing catch phrase.