作者: Mehul Motani , Chandra R Murthy , Rajshekhar Vishweshwar Bhat , Rahul Vaze
DOI:
关键词:
摘要: We consider an energy harvesting transmitter equipped with two batteries having finite storage capacities, communicating over additive white Gaussian channel. The work is motivated by observation that many practical batteries, when repeatedly charged after being partially discharged, suffer from degradation in the usable capacity. capacity can be recovered completely discharging battery before charging it fully again. Hence, this work, we impose constraint a must (discharged) only discharged (charged). Our goal to maximize longterm average throughput non-causal and causal knowledge of arrivals, which assume Bernoulli. propose sub-optimal policies obtain upper bound on performance gap (G) optimal long-term achieved infinite batteries. find G remains constant as amount harvested per arrival increases. Numerically, also decreases faster than inverse square root for specific parameters.