Error Resilience Algorithm for Haptic Communication Based on Remedy-LSTM

As a new type of immersion interaction method, haptic communication technology has been widely applied in various fields. Data loss is inevitable during haptic communication, which will have significant negative impact on user's experience. Error resilience algorithm (ERA) is an effective method to solve this problem. However, traditional ERAs are based on linear prediction methods. Existing studies have verified that haptic data is not linear. Therefore, there still leave gaps to improve the performance of ERAs for haptic communication. To this end, this paper proposes an ERA of haptic communication based on an improved long short-term memory (LSTM) neural network. Firstly, an improved LSTM network is constructed by adding remedy gates to realize haptic data prediction, which effectively reduces the prediction error. Then, the presented ERA is implemented with the prediction model. Finally, we establish a simulation platform to compare the performance of the proposed algorithm with the popular-used ERAs in haptic communication. Experimental results show that our algorithm.