Base Station Sleep Modes to Trade-off Energy Saving and Performance in 5G Networks.

With the dynamic growth in ICT industry from recent years, energy demand is increasing with the demand of high data rates in mobile cellular communication. Base Stations(BSs) contribute to 80% of energy consumption of overall mobile cellular radio network. Efforts are reinforced for energy efficient system design for cellular networks concerning worldwide increase in energy costs and CO2 emissions impacting ecological balance. Under the energy efficiency incorporation studies for BSs, Sleep Modes(SMs) are discussed widely to save energy consumption by radio resources in low utilization periods by turning them off. In this thesis, a simulation study is carried out by using Advance Sleep Modes(ASMs)[1] for 5G network New Radio (NR) signaling and other IOT applications. Sleep modes are composed of different duration and have specific switching time. Hence, according to arrival rates and service time, BS can benefit from SMs by achieving lower energy consumption with respect to standard setting where BS is always active. Consequently, a reactivation delay adds in the system and as 5G networks are thought to be ultra-responsive with low latency we need to achieve energy saving keeping the constraint of reliability and QOS guarantees. Therefore, additional delays added due to ASMs are investigated and reasoned after reactivation time required for arrivals happened during BS is in sleep mode. A formulation of sleep mode strategy is considered where an adjustment is made in minimum sleep mode durations for each SM which resulted in a tradeoff between latency added due to reactivation and energy saving. To better look at realistic traffic scenarios, considering a given prediction for weekdays and weekend arrival patterns for both Residential and Metropolitan areas, these traffic patterns are investigated deeply by simulating BS under ASMs and useful findings are achieved regarding overall energy consumption during the day. Whereby, the effect of error in prediction is also discussed by looking at delays and energy savings using the same parameter settings. In the end, a detail overview is presented which includes the analysis of efficient ways of using ASMs for concerning energy costs for Mobile Operators as well as keeping service level constraints regarding 5G mobile radio network into consideration.