xApp for traffic steering and load balancing in the O-RAN architecture

Traffic steering is an essential aspect of the radio access network (RAN). The O-RAN Alliance architectural framework provides an environment for intelligent traffic steering using different AI/ML techniques. In this work, we present an xApp in the RAN intelligence controller (RIC) for traffic steering and load balancing to ensure that the user equipment (UE) achieves an acceptable throughput. We present a K-means learning clustering technique on the UEs based on the throughput and quality of service metrics. The clustering technique is used to determine UEs experiencing low throughput for handover. Cell throughput prediction is then performed using long-short-term memory (LSTM) to predict the average cell throughput generated by the individual cells. A steering algorithm is developed to select UEs for handover. A Handover request message is generated and then sent to the E2 nodes. The achieved results demonstrate the effectiveness of our proposed algorithm with a significant gain in throughput and a fair distribution of UEs among cells.