Shared Buffer for OTTT VOD Retransmission

The patent merges per-client buffering into a single shared buffer that serves all clients. Instead of maintaining separate buffer instances for each client connection, the system uses one common buffer to store outgoing data packets that may need retransmission. This consolidation eliminates the linear scaling of memory consumption with client count while preserving retransmission capability through centralized packet storage and selective retransmission based on client-specific ACK/NACK feedback.

The patent segments the data handling process into shared data storage and client-specific operations. The actual data packets are stored once in the shared buffer, while client-specific operations (such as generating individual ACK/NACK packets and selecting which packets to retransmit to which client) are performed separately. This segmentation eliminates redundant copying of data while maintaining proper per-client data management through selective packet retrieval and customization.

The patent combines multiple per-client buffer requirements into a single shared buffer, dramatically reducing the total memory capacity needed for retransmission handling. Instead of allocating separate memory regions for each client connection, the system uses one pooled buffer that serves all clients. This merging reduces both the absolute memory quantity required and the CPU memory bandwidth consumption, as the CPU accesses a single buffer structure rather than traversing multiple client-specific buffer hierarchies.