HTTP ProSe Polling Scheduling for Reduced Latency
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current HTTP-based ProSe discovery mechanisms in LTE systems face challenges with delivery delays and overhead due to polling intervals and long polling schemes, which are inefficient for proximity alert procedures and UE location reporting.
Innovation Solution
Implementing a method that allows UEs to choose polling times based on location reporting schedules, enabling the network to determine when to initiate Proximity Alerts or Cancel Proximity Requests, and adjusting reporting schedules for more frequent updates when UEs are likely to be in proximity, thus optimizing HTTP connection usage.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If polling procedure is used to enable server-initiated transactions in HTTP-based ProSe discovery, then the system can support proximity alert procedures, but delivery delay increases due to inter-polling intervals
Solution Approach 1:
Instead of the client polling the server at fixed intervals to check for data, the invention inverts the approach by allowing the server to push data to the client immediately when available, eliminating the waiting period inherent in traditional polling. This is achieved through a modified HTTP mechanism where the server can initiate data transmission without requiring a client request, thus resolving the contradiction between supporting server-initiated transactions and minimizing delivery delay.
2Loss of time
If long polling is used to reduce delivery latency, then data can be delivered immediately when available, but HTTP connection overhead increases significantly
Solution Approach 1:
The invention implements a periodic action principle by establishing HTTP connections at regular intervals rather than maintaining continuous long-polling connections. The client periodically initiates HTTP requests to check for data, and the server responds with data if available or with an indication that no data is present. This periodic approach achieves timely data delivery while allowing connections to be closed between periods, thus reducing the overhead associated with maintaining persistent connections.
3Speed
If long polling keeps HTTP connection open to reduce latency, then delivery speed improves, but blocking of new outgoing requests occurs
Solution Approach 1:
The invention segments the HTTP communication into discrete, independent request-response cycles rather than maintaining a single prolonged connection. Each HTTP request is a self-contained unit that completes quickly with a clear response, allowing the client to immediately initiate the next request if needed. This segmentation eliminates the blocking effect where one long-polling connection would prevent other requests from being sent, while still achieving fast data delivery through efficient use of the HTTP protocol.
4Reliability
If HTTP server maintains active connection with each client UE for long polling, then real-time data delivery is possible, but network overhead increases with number of clients
Solution Approach 1:
The invention implements self-service by having each client autonomously manage its own HTTP connections and polling schedule without requiring the server to maintain persistent connections. Clients independently initiate requests when they need to check for data, and the server simply responds to these requests without needing to actively manage or maintain connections. This shifts the connection management burden from the server to the clients, allowing the server to handle multiple clients efficiently without proportionally increasing overhead, while still enabling real-time data delivery when available.
Data Source
Figure 1
Figure 2
Figure 3
AI summary
Example methods and apparatuses for managing polling in devices implementing proximity services are presented. For instance, an example method of polling management in a ProSe system is presented, which includes receiving, at a network entity, a polling message from a first UE. In addition, the example method may also include receiving, at the network entity and after receiving the polling message, one or both of a first location report associated with the first UE and a second location report associated with a second UE. Furthermore, the example method may include determining whether to generate a polling response message upon receiving one or both of the first location report and the second location report, wherein the polling response message includes a next polling time for the first UE that is based on a location reporting schedule associated with one or both of the first UE and the second UE.