September 02, 2019 by Luke James
The intensive use of Wi-Fi communications has led to the growth of W-Fi infrastructure worldwide. Due to the huge amounts of data transmission, however, these consume a whole lot of power that is lost as heat. This shouldn't (and doesn't need to) be the case.
Wi-Fi is great, but it's a major hog when it comes to energy consumption. That's simply the price we pay for having Wi-Fi though, right? Well, not exactly. In fact, only a small amount of energy burned by Wi-Fi infrastructure is used to supply the systems themselves; most of it is lost through heat because Wi-Fi systems and infrastructure have terrible overall efficiency.
The success of wireless communications worldwide has led to a massive expansion in infrastructure. Base stations and cell towers are popping up everywhere, but not enough is being done to tackle efficiency.
The knock-on effect of this is not only higher operational costs (and thus higher consumer costs) but increased greenhouse gas emissions, too.
A telecommunications tower that can be used to receive and transmit WiFi signals. Image courtesy of Pixabay.
We have known for years just how much of an energy hog Wi-Fi infrastructure is. It has been covered widely in the media and several attempts to mitigate the problem have been theorized but little has come of them.
The most promising step towards more efficient Wi-Fi thus far has been so-called 'Passive Wi-Fi' which is up to 10,000 times more power-efficient.
To achieve this, a team of electrical engineers and computer scientists at the University of Washington detached the digital and analog operations involved in Wi-Fi radio transmissions, the latter of which being the power-consuming monster that makes Wi-Fi such a bad waster of energy.
Passive Wi-Fi's architecture separates the power-intensive analog functions and the power-efficient digital functions and assigns them to a single network device that is plugged into a wall socket. A range of sensors then produce Wi-Fi packets using little power by reflecting and then absorbing the signals from the network device.
The result is that the plugged-in device does all the hard work but only when reflecting signals, thus reducing power consumption.
A repairman working on a cell tower. Image courtesy of Pixabay.
Cost is one thing; damage to the environment is another.
A real-world problem of low power efficiency in wireless operations is the huge amount of heat that is lost. This is a problem for companies that must foot the bill, but also for wider society who must deal with the consequences of rising greenhouse gas emissions.
It is these concerns that motivated recent research by an MIT PhD candidate, Omer Tanovic, from the Department of Electrical Engineering and Computer Science. Tanovic joined the Laboratory for Information and Decision Systems (LIDS) to work on the design of signal processing systems that will improve power efficiency.
Tanovic's latest project hopes to address power efficiency problems by decreasing the peak-to-average power ratio (PAPR) of wireless communication signals. The lower the PAPR ratio, the more efficient the wireless transmission. That is because PAPR is an indicator, albeit indirectly, of how much power is needed to send and receive a clear signal across a wireless network.
To do this, he plans to use mathematics and develop algorithms that will guarantee improved system performance. Tanovic has already designed an algorithm that can reduce the PAPR of current communications signals, allowing systems to operate closer to maximum efficiency.
“Every cell tower has to have some kind of PAPR reduction algorithm in place in order to operate. But the algorithms they use are developed with little or no guarantees on improving system performance,” Omer says. “A common conception is that optimal algorithms, which would certainly improve system performance, are either too expensive to implement — in terms of power or computational capacity — or cannot be implemented at all.”
The biggest consumer of energy in a cellular network is the power amplifier. This converts low-power signals into higher-power output. It does this because signals need to have a signal-to-noise ratio strong enough to reach its destination, e.g. a cell tower.
Currently, the issue with power amplifiers is that they are designed in a way that allows them to carry signals with peaks much, much higher than that of the signal power being transmitted. This means that they are often operationally inefficient.
Tanovic's algorithm, designed to decrease the PAPR of signals and thus enabling power amplifiers to operate at maximum efficiency, is just one possible solution to this problem.
Power amplifiers are difficult to design though. Rather than starting fresh, perhaps mathematical processes are the best way to tackle Wi-Fi inefficiency, at least in the short-term.
