A good overview

For those readers who need to understand the RTP protocol, either because of network management or from the standpoint of inventing algorithms for playout, this book is suitable. The author gives mostly a descriptive overview of the protocol, but he does include brief discussions on some of the quantitative aspects behind it, such as how to calculate jitter and compensation for clock skew.
For those readers interested in RTP performance issues, some of the topics of interest that are not discussed effectively in the book include:
- Case studies of the affect of RTCP packets on network congestion. For example, studies on the dependence of network congestion (if any) on the reporting interval. The author makes the statement to the effect that an incorrect implementation of RTCP will result in a linear dependence on the number of users and cause "significant" network congestion. This assertion needs justification either through modeling or actual testing.
- Along these same lines, more discussion on how step joining can congest the network and how updating the number of senders can alleviate this congestion.
- The possibility that RTCP may act as a competing risk to some admission-control implementations. In particular, the feedback obtained by receiver reports allowing senders to adapt their transmissions is very similar to what is done using measurement-based policy-based admission control (MBAC).
- In-depth discussion on the optimal round-trip time for interactive applications.
- Case studies illustrating the difficulties in generating compound RTCP packets when there are more than 31 active senders.
- Examples or case studies showing how a poor choice of hash function for obtaining the synchronization source identifier (SSRC) can lead to "unbalanced and inefficient operation."
- The effect on congestion (if any) of the overhead of marked packets for informing receivers when to perform join experiments in multicast congestion control.

In spite of these shortcomings, there are a lot of interesting discussions in this book, such as the use of white noise and pattern-matching to conceal audio losses, the Rate Adaptation Protocol (RAP), and the use of layered coding for congestion control in multicast. Readers who need more details can consult the references given at the end of the book.

Note: This review is based on a reading of chapters 1-11 of the book.

Other Reviews

• A good overview
• Excellent ouvrage sur le sujet
• comprehensive explanation
• a must-have book if you are in this area
• Excellent material - well presented