Blind Detection and Prediction of Multi-SIM UE Subframe Loss

The share of mobile User Equipments (UEs) supporting insertion of Multiple Subscriber Identity Modules (Multi-SIM) has increased significantly throughout the world. Due to cost optimization, Multi-SIM UEs are often designed such that they discard chunks of data in one subscriptions active data stream in order to support network communication for another subscription, due to radio hardware or software access conflicts. Limited signalling between the UE and Base Station (BS) makes it such that the UE has no way to signal that it is a Multi-SIM UE, or that it will discard data. We present an algorithm based on Markov chains which is able to blindly detect if a connected UE is a Multi-SIM UE and predict potential data loss. During operation, the state and transition models used in the algorithm are continuously updated in order to let the algorithm adapt to the pattern of discarded data in the active connection. The algorithm is investigated within the KPIs of; number of correct detections made, number of missed gaps, and number of false predictions. The analysis is created using a link level simulator and the presented results show that the algorithm is able to detect and predict Multi-SIM gaps, and as the algorithm learns it adapts the prediction to the actual operations of the UE.