Secure indoor traffic modeling in long‐term evolution femtocell networks using efficient Markov decision process

Mobile users spent most of their time in an indoor environment, therefore it is crucial to develop indoor traffic models for next‐generation networks. The significance of this paper is related to the predominant femtocell technology as it is anticipated to be a very crucial part of the modern mobile networks such as 5G and beyond networks. The Femtocell network can satisfy the increased bandwidth demand for video and data traffic with improved network capacity and security. Indoor data traffic exhibits different characteristics that need to be dealt with distinct criteria of the Poisson process and existing classifying techniques. Indoor traffic models, proposed in the literature do not consider long‐range dependent characteristics for traffic modeling. Moreover, these traffic models have the ability to capture large flows only. In this paper, we have provided an efficient quality of service‐based classifier for downlink transmission in long‐term evolution (LTE)‐femtocell network. Our proposed algorithm models the indoor traffic by Poisson Lomax Burst Process. Modeled traffic streams, after prioritization, are classified using Markov Decision Process (MDP) classifier. Furthermore, a detailed comparison of our proposed MDP algorithm with the state‐of‐the‐art algorithms in the literature has also been presented. The network performance parameters have been assessed in terms of throughput, delay, and jitter. Simulation results depict a significant improvement in throughput in contrast with Round Robin (RR), which is approximately 8%, 6%, 6%, and 7% for voice, video, BE, and BG services, respectively, while modified least weighted delay first (MLWDF) gives least performance among them. MDP also provides an improvement in delay and average jitter of four different traffic flows and mixed traffic users when compared with RR and MLWDF algorithms. Long‐term evolution femtocell architecture.