Research And Implementation Of Edge Caching Optimization For XR Extrasensory Experience In Smart Tourism

With the support of 5G communication technology,the domestic tourism industry has achieved a deeper level of development,gradually moving from intelligent tourism to a new stage of intelligent cultural tourism.Meanwhile,new technologies such as Extended Reality(XR)technology,human-computer interaction technology and artificial intelligence technology break through the limitations of the original tour,bringing a more amazing cultural and tourism experience.However,the development of new services has brought a great data transmission processing load,especially XR technology brings interactive and immersive super-sensory experience for visitors while the data it requires is continuous in 360° space and occupies great storage,and the traditional caching method is difficult to meet the demand.Edge caching is also a new feature brought by 5G technology,which provides distributed caching services for visitors who are also at the same edge by using gateways,routing,and visitor devices at the edge of the network,with the goal of reducing the network congestion and the response times in the central cloud data centers and enabling a caching experience with low-latency,high-bandwidth,real-time aware,security-enhanced.This thesis is a study of edge caching strategies for smart scenic scenarios,centered on the new service requirements of XR.Since the XR data is difficult to cache directly due to the limited edge cache space,the XR data is processed and prioritized in blocks by a grid-based segmentation method before caching.Then,according to the tourist trajectory data,the LSTM neural network is used to predict the attractions to be visited by visitors.After the prediction results are obtained,Q-Learning integrates the cache status and prediction accuracy of each edge gateway to select the appropriate cache size to pre-cache the XR data related to the predicted attractions,thus reducing the time spent on data transmission process.Simulation results show that the algorithm enables efficient caching of XR data on memory-limited edge caches,reducing the time used for transmission and improving the visitors' experience of using new XR services.Based on the construction of edge caching strategy,this thesis designs and develops a smart scenic edge caching distribution system on PyCharm platform using Python language.The system consists of a request prediction module and a cache decision module,in which the prediction module uses historical visitor trajectory information to predict the distribution of requests,and then the cache decision module uses the prediction results to achieve cache distribution at the edge gateway,improving the hit rate and utilization rate of the cache and providing a strong impetus to the application process of new services in the smart scenic area.