Channel-aware multi-level NOMA-ALOHA game

Abstract

Random access (RA) has been widely used as a critical technique for the medium access control layer of the Internet of Things applications. Slotted ALOHA, one of the RA protocols, uniquely offers the benefits of minimal complexity and the removal of partially overlapping transmissions, lowering the frequency of collisions. But as the traffic load and the number of devices grows, it may experience congestion. To this end, we apply the Nonothogonal multiple access technique to ALOHA. Considering the non-cooperative behavior of the Aloha network, we set up a NOMA-ALOHA game. We use the power efficiency of each user in the instantaneous channel to establish the payoff function and find the Nash equilibrium. The simulation results show that this kind of game can reduce user conflicts and improve the utilization of network resources. Furthermore, the increasing the number of power levels offered, the higher the system’s throughput and the shorter the time it takes to send a packet. In addition, we use the location information of each user to estimate the transmitting power of each user based on path loss. Therefore, we propose an indoor positioning system based on sensed Received Signal Strength (RSS) values and the Adaboost algorithm. In order to delete suspicious data and create a fingerprint database, the method employs the RSS threshold value process. The sample weights of the training set are modified continuously to prepare for classification using the Adaboost algorithm’s continuous iteration. The experimental findings indicate that the lower the RSS threshold, the lower the positioning error rate. Furthermore, the system is general, so there is no need to model each unit individually. In comparison to the current system proposed in the literature, the proposed process will accommodate a larger number of access points with better positioning accuracy.