Towards greener good cities with machine learning-based ‘sleep schedules’

The idea of good cities is based on subtle mobile networks that will not solely join people sooner or later but in addition people to different good units. However, this may additionally require large vitality consumption. In the wake of local weather change, this will make issues worse for the environment by growing the greenhouse gasoline emissions. Thus, we not solely want good cities but in addition greener good cities.

One approach to handle this problem is by switching off base stations (BSs), radio transmitters/receivers that function the hub of the native wi-fi community, once they have little to no site visitors load. Laboratory testing has proven that lively BSs devour as a lot as 60% of the utmost energy consumption even underneath no site visitors load and switching them off can carry it right down to 40%. However, there’s a trade-off: placing BSs to sleep makes their site visitors logs unavailable, which additionally reduces the accuracy of site visitors prediction. Is there a approach to keep away from this compromise between accuracy and sustainability?

The reply, in response to a brand new examine, appears to be “yes.” The examine, led by Professor Ryoichi Shinkuma from Shibaura Institute of Technology (SIT), Japan, and his colleagues, Associate Professor Kaoru Ota from Muroran Institute of Technology, Japan and Associate Professor Takehiro Sato from Kyoto University, Japan, proposed a novel scheme that not solely decreased vitality consumption however demonstrated a better site visitors prediction accuracy in comparison with the benchmark schemes. This paper was revealed within the journal IEEE Network Magazine in November/December 2021.

How did the researchers obtain this outstanding feat? Prof. Shinkuma explains, “We applied software defined network (SDN) and edge computing to a cellular network such that each BS is equipped with an SDN switch, and an SDN controller can turn off any BS according to the traffic prediction results. An edge server collects the traffic logs through the SDN switches and predicts traffic volume using machine learning (ML).”

The ML methodology utilized by the researchers determined which BSs may very well be put into “sleep mode” primarily based on the significance of their site visitors logs in enhancing the prediction accuracy. Thus, BSs with low contribution to the accuracy for earlier time slots have been put to sleep on the subsequent slot to avoid wasting vitality.

To validate their scheme, the researchers used real-world cellular traffic knowledge collected over two months and in contrast its efficiency towards that of two benchmark schemes. To their delight, the brand new scheme outperformed the benchmark schemes in its robustness towards lowering the variety of lively BSs and totally different BS units.

Could this examine be a harbinger of greener cellular networks and good cities? Prof. Shinkuma is optimistic. “By intelligently controlling the operation of BSs, renewable energy sources could be used to power future networks, and depending on the availability of renewable energy resource, the sleep schedules of the BSs can be determined,” he speculates.

Will machine studying assist us go inexperienced in our quest for good cities? We can not wait to see.

Provided by
Shibaura Institute of Technology