MPEG-TS Packet Loss Detection via Counter Segmentation
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Conventional continuity counters in communications networks fail to detect packet loss when the number of lost packets is a multiple of the values they can represent, leading to undetectable losses.
Innovation Solution
The implementation of encapsulating packets that encapsulate a number of original streaming packets in such a way that the number of encapsulated packets lacks common factors with the number of values representable by the continuity counter, allowing for the detection of packet loss through gaps in continuity counter values and unusual inter-packet arrival times.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If conventional continuity counters are used to detect packet loss, then packet loss detection is possible for most cases, but packet loss remains undetectable when the number of lost packets is a multiple of the values representable by the continuity counter
Solution Approach 1:
The transmitter segments the original stream of packets into multiple sub-streams, each with its own continuity counter. By distributing packets across multiple sub-streams and transmitting them in interleaved fashion, the system ensures that if a multiple of N packets are lost from the original stream, at least one sub-stream will show a detectable gap in its continuity counter values, thereby resolving the undetectable packet loss problem.
2Ease of operation
If the continuity counter uses a finite number of values (e.g., 2-bit counter with 4 values), then the identifier sequence can be implemented, but the counter rolls over and creates gaps that mask packet loss when the loss is a multiple of the counter's range
Solution Approach 1:
The system divides the packet stream into multiple sub-streams, each using its own continuity counter with finite values. This segmentation allows the system to maintain simple finite counters while improving reliability through the mathematical property that at least one sub-stream will exhibit a detectable gap when packets are lost in multiples of the counter range.
Solution Approach 2:
The invention adds a new dimension to packet identification by creating multiple parallel continuity counters instead of relying on a single counter. This dimensional expansion from one counter to multiple counters allows the system to detect packet loss patterns that would be invisible to a single counter, effectively adding redundancy in the detection space.
Data Source
AI summary
Methods, mediums, and systems are described for detecting the loss of one or more packets in a data stream and identifying the missing packets so that they may be re-transmitted. The packets of the data stream may be associated with a continuity counter capable of representing a number n of values. A number m of packets of the data stream may be encapsulated into an encapsulating packet, where m does not share any common factors with n. Accordingly, if a number of streaming packets are lost, the loss of the packets may be readily detected even when the number of lost packets is a multiple of n. The received packets may be searched for a range including the missing packets and bounded by packets that are likely to be unique. Once the bounding packets are identified, packets between the bounding packets may be requested from the original sender.