Providing high quality live video over wireless networks is a challenging problem due to the lossy wireless channels and the large bandwidth requirements of video traffic. Within this large problem space, we focus on the efficient broadcast of a scalable video over wireless PMP networks using network coding. Such a scalable video imposes an order constraint on decoding the layers and has a hard deadline before which the video layers need to be decoded [13]. Therefore, using the scarce radio resources, as many video layers as possible need to be in-order delivered to the viewers before the deadline. In fact, the video coding community advocates for taking into account the interdependence of video layers and the hard deadline in designing a video-aware transmission scheme. On the other hand, network coding community has well established that network coding offers an increased throughput, simple scheduling, delay reduction and enhanced security in wireless PMP networks [40–43, 52]. This chapter bridges the gap between these two techniques and develops a unified video-aware network coded scheme that improves both video quality and throughput.
Our work of this chapter is inspired by the recent works on scalable video transmission using RLNC in [17, 43, 44, 97]. In this chapter, we adopt XOR based IDNC to investigate its performance for scalable video transmission. In addition to the inherent encoding and decoding simplicity, the throughput performance of IDNC closely follows that of RLNC in a
network with a small number of devices [93]. In particular, IDNC schemes can achieve the optimal throughput for the two-device or three-device system as shown in [53,70]. Moreover, as video streaming applications continue to proliferate, the wireless sender (e.g., the base station) often needs to support multiple simultaneously running applications with heteroge- nous video characteristics. Therefore, the sender can adopt IDNC to encode packets from different video sequences together. This allows immediate decoding of the received packets of different video sequences and immediate use of decoded packets at the applications, es- pecially when one or more video sequences are encoded using multiple description coding in addition to scalable video coding.
In this chapter, we are interested in designing an efficient IDNC framework that maxi- mizes the minimum number of decoded video layers over all devices before the deadline (i.e., improves fairness in terms of the minimum video quality across all devices). We consider that a service provider adopts a maxmin policy to improve fairness across all devices regardless of their channel conditions. With such a policy, some devices experiencing harsh channel conditions are prioritized over other devices experiencing good channel conditions with the aim of delivering an acceptable video quality to all devices. This may prevent a severe degradation of the quality of services at a device experiencing a poor channel condition.
For such scenarios, by taking into account the deadline, the coding decisions need to carefully balance coding only from the base layer versus coding from all video layers. While the former guarantees the highest level of priority to the base layer, the latter increases the possibility of decoding a large number of video layers before the deadline. In this context, our main contributions are as follows:
• We derive an upper bound on the probability that a given number of video layers are decoded by all devices before the deadline. Using this probability, we are able to approximately determine whether the broadcast of a given number of video layers can be completed before the deadline with a predefined probability.
• We design two prioritized IDNC algorithms for scalable video, namely the expanding window IDNC (EW-IDNC) algorithm and the non-overlapping window IDNC (NOW- IDNC) algorithm. EW-IDNC algorithm selects a packet combination over the first video layer and computes the resulting upper bound on the probability that the broad- cast of that video layer can be completed before the deadline. Only when this proba-
bility meets a predefined threshold, the algorithm considers additional video layers in coding decisions in order to increase the number of decoded video layers at the devices. In contrast, NOW-IDNC algorithm always selects a packet combination over the first video layer without exploiting the coding opportunities by including additional video layers.
• We use a real scalable video sequence to evaluate the performance of our proposed algorithms. Simulation results show that our proposed EW-IDNC and NOW-IDNC algorithms increase the minimum number of decoded video layers over all devices compared to the IDNC algorithms in [9, 52] that are oblivious to the in-order layer decoding and deadline constraints. Furthermore, the proposed algorithms also achieve a comparable performance compared to the expanding window RLNC algorithm in [43, 44] while preserving the benefits of IDNC strategies. It is also observed that EW- IDNC achieves a similar performance in terms of maximizing the minimum number of decoded video layers and significantly better performance in terms of maximizing the mean decoded video layers as compared to NOW-IDNC algorithm.
The rest of this chapter is organized as follows. The system model and IDNC graph are described in Section 3.2. We illustrate the importance of appropriately choosing a coding window in Section 3.3 and draw several guidelines for prioritized IDNC algorithms in Section 3.4. Using these guidelines, we design two prioritized IDNC algorithms in Section 3.5. We formulate the problem of finding an appropriate packet combination in Section 3.6 and design a heuristic packet selection algorithm in Section 3.7. Simulation results are presented in Section 3.8. Finally, Section 3.9 concludes the chapter.