The Fifth Generation (5G) mobile networks use millimeter waves (mmWaves) to offer gigabit data rates. However, unlike microwaves, mmWave links are prone to user and topographic dynamics. They easily get blocked and end up forming irregular cell patterns for 5G. This in turn causes too early, too late, or wrong handoffs (HOs). To mitigate HO challenges, sustain connectivity, and avert unnecessary HO, we propose an HO scheme based on a jump Markov linear system (JMLS) and deep reinforcement learning (DRL). JMLS is widely known to account for abrupt changes in system dynamics. DRL likewise emerges as an artificial intelligence technique for learning highly dimensional and time-varying behaviors. We combine the two techniques to account for time-varying, abrupt, and irregular changes in mmWave link behavior by predicting likely deterioration patterns of target links. The prediction is optimized by meta training techniques that also reduce training sample size. Thus, the JMLS-DRL platform formulates intelligent and versatile HO policies for 5G. When compared to a signal and interference noise ratio (SINR) and DRL-based HO scheme, our HO scheme becomes more reliable in selecting reliable target links. In particular, our proposed scheme is able to reduce wasteful HO to less than 5% within 200 training episodes compared to the DRL-based HO scheme that needs more than 200 training episodes to get to less than 5%. It supports longer dew time between HOs and high sum rates by ably averting unnecessary HOs with almost half the HOs compared to a DRL-based HO scheme.
Keywords: Fifth Generation; deep reinforcement learning; handover; jump Markov linear system; millimeter bands.