Layered Source Data FEC for Mixed-QoS Packet Delivery
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
MPEG-2 TS faces challenges in supporting multimedia services due to unidirectional communication, transmission inefficiency, and overhead issues, particularly in hybrid content delivery services where different types of content have varying Quality of Service (QoS) requirements, leading to data loss and quality degradation.
Innovation Solution
A two-stage FEC encoding scheme is implemented to encode source blocks containing different types of data with varying QoS requirements, where the first stage encodes sub-blocks and the second stage encodes the entire source block or specific data segments, with coded signal information including position and type information for efficient error correction.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Adaptability or versatility
If a single FEC encoding scheme is used for all data types in hybrid content delivery, then the system complexity is reduced, but the QoS requirements of different data types (audio, video, file) cannot be met simultaneously, leading to quality degradation
Solution Approach 1:
The source block is divided into multiple sub-blocks, where each sub-block contains data of the same type (audio, video, or file). This segmentation allows different FEC encoding schemes to be applied to different sub-blocks according to their specific QoS requirements, thereby achieving adaptability without overwhelming complexity.
Solution Approach 2:
Different FEC encoding parameters (such as code rate, block size, or error correction strength) are applied locally to each sub-block based on its data type and QoS requirements. For example, file data sub-blocks may use stronger error correction while video sub-blocks use optimized parameters for real-time playback, achieving local optimization of quality.
2Adaptability or versatility
If data is transmitted in packets over heterogeneous networks, then network flexibility and IP compatibility are improved, but data loss occurs in packet units causing quality degradation and service failure
Solution Approach 1:
Forward error correction codes are added to the data blocks before transmission over the packet-switched network. This preliminary protective action enables the receiver to recover from packet losses without requiring retransmission, maintaining reliability while preserving network flexibility and IP compatibility.
Solution Approach 2:
Redundant error correction data is embedded in advance into the transmitted packets, creating a cushion against potential packet loss. When packets are lost during transmission over heterogeneous networks, the receiver can use the pre-added error correction information to reconstruct the original data, preventing quality degradation.
3Reliability
If MPEG-2 TS is used for broadcast transmission, then error resilience in erroneous transmit environments is improved, but transmission efficiency decreases due to fixed frame size and unnecessary overhead
Solution Approach 1:
The system transitions from the fixed frame structure of MPEG-2 TS to dynamic packet sizes in the MMT framework. Data blocks are segmented into packets with flexible sizes adapted to the actual content and QoS requirements, eliminating the waste associated with fixed frame sizes while maintaining error resilience through adaptive error correction.
Solution Approach 2:
The packetization parameters (such as packet size, segmentation depth, and error correction code rate) are changed and optimized according to the specific transmission environment and content type. This allows the system to achieve both error resilience and transmission efficiency by adjusting parameters dynamically rather than being constrained by fixed MPEG-2 TS parameters.
Data Source
Figure 1
Figure 2a
Figure 2b
AI summary
An encoding apparatus and a method of encoding a source block including different types of data payloads are provided. The method includes dividing the source block into a predetermined number M of sub blocks, generating a predetermined number PI of base parity payloads that correspond to each of the predetermined number M of sub blocks by performing first Forward Error Correction (EEC) encoding on each of the predetermined number M of sub blocks, generating a predetermined number P2 of extended parity payloads that correspond to the source block by performing second PEC encoding on a particular type of data payloads among data payloads that make up the source block, and configuring a source coded block based on a predetermined number N of sub coded blocks including the predetermined number M of sub blocks and the predetermined number PI of base parity payloads generated.