Kerim Fouli

Network coding in next-generation passive optical networks

As the emerging access architecture, NGPONs feature enhanced PON configurations, metro-access integration, and bimodal fiberwireless (FiWi) networks. The moving landscape of NG-PONs provides opportunities for applying novel and promising technologies such as network coding (NC). In this work, we introduce the basic principles of NC and discuss their applicability to NG-PONs, with a focus on layer 2 design. Our example illustrations and simulations demonstrate significant potential performance improvements in various NG-PON scenarios while clarifying some underlying topological constraints of NC.

Network Coding as a WiMAX Link Reliability Mechanism

We design and implement a network-coding-enabled reliability architecture for next generation wireless networks. Our network coding (NC) architecture uses a flexible thread-based design, with each encoder-decoder instance applying systematic intra-session random linear network coding as a packet erasure code at the IP layer. Using GENI WiMAX platforms, a series of point-to-point transmission experiments were conducted to compare the performance of the NC architecture to that of the Automatic Repeated reQuest (ARQ) and Hybrid ARQ (HARQ) mechanisms. In our scenarios, the proposed architecture is able to decrease packet loss from around 11-32% to nearly 0%; compared to HARQ and joint HARQ/ARQ mechanisms, the NC architecture offers up to 5.9 times gain in throughput and 5.5 times reduction in end-to-end file transfer delay. By establishing NC as a potential substitute for HARQ/ARQ, our experiments offer important insights into cross-layer designs of next generation wireless networks.

Low-Latency Polling Schemes for Long-Reach Passive Optical Networks

The increased propagation delay of future long-reach passive optical networks (LR-PONs) may lead to a significantly increased idle time and delay if optical network units (ONUs) use conventional report-grant mechanisms. Sophisticated and efficient bandwidth allocation mechanisms are required to cope with the imposed propagation delay in LR-PONs. In this study, we evaluate three dynamic bandwidth allocation (DBA) frameworks in terms of frame (packet) delay; namely, we consider conventional (interleaved) polling for traditional PON and two recently introduced scheduling paradigms for next generation LR-PON, i.e., multi-thread polling (MT-P) and real-time polling (RT-P). We enhance MT-P and RT-P by applying the just-in-time framework. Next, we provide an analytical framework for evaluating the end-to-end frame delay in our enhanced MT-P and RT-P frameworks. We compare their performance with conventional polling and double-phase polling and investigate their shortcomings and advantages in an LR-PON setting. The simulation results closely match the analysis for this framework. Also, our results indicate that RT-P significantly reduces frame delay in LR-PONs compared to MT-P and conventional polling frameworks.

Congestion control for coded transport layers

The application of congestion control can have a significant detriment to the quality of service experienced at higher layers, especially under high packet loss rates. The effects of throughput loss due to the congestion control misinterpreting packet losses in poor channels is further compounded for applications such as HTTP and video leading to a significant decrease in the users quality of service. Therefore, we consider the application of congestion control to transport layer packet streams that use error-correction coding in order to recover from packet losses. We introduce a modified AIMD approach, develop an approximate mathematic model suited to performance analysis, and present extensive experimental measurements in both the lab and the “wild” to evaluate performance. Our measurements highlight the potential for remarkable performance gains, in terms of throughput and upper layer quality of service, when using coded transports.

Network Coding as a WiMAX Link Reliability Mechanism: An Experimental Demonstration

Our demonstration showcases a network-coding (NC)– enabled reliability architecture for next generation wireless networks. Our NC architecture uses a flexible thread-based design, applying systematic intra-session random linear network coding as a packet erasure code at the IP layer. Using GENI WiMAX platforms, a series of point-to-point transmission experiments are conducted to compare NC with Automatic Repeated reQuest (ARQ) and Hybrid ARQ (HARQ). At the application layer, Iperf and UFTP are used to measure throughput, packet loss and file transfer delay. In our selected scenarios, NC offers up to 5.9 times gain in throughput and 5.5 times reduction in file transfer delay, compared to HARQ and joint HARQ/ARQ. Our demonstration hence illustrates that lower-layer redundancy mechanisms such as HARQ and ARQ incur high cost since they operate at the packet-level. Conversely, running NC at higher layers (e.g., IP) amortizes the cost of redundancy over several packets, thus leading to higher efficiency.

Novel FWM-Based Spectral Amplitude Code Label Recognition for Optical Packet-Switched Networks

We propose and demonstrate a novel architecture for four-wave mixing (FWM)-based recognition of spectral amplitude code (SAC) labels in optical packet-switched networks. With a proper code design, a unique FWM idler for each SAC label, referred to as a label identifier (LI), is generated in a nonlinear medium. A serial array of fiber Bragg gratings is then used to reflect the LI wavelengths. Each LI is associated with a unique amount of delay between two optical signals received at two photodiodes (PDs). Label recognition is then achieved by measuring this unique time delay (referred to as the characteristic delay). The main advantages of the proposed method include the following: no serial-to-parallel conversion is required, simple label extraction is achieved, variable-length packets are supported, and the number of PDs used in the label recognition module is reduced. Moreover, the LI wavelengths do not need to exhibit any periodicity or match a particular wavelength grid; this results in a less challenging code design with smaller spectral occupancy for label generation. An experiment is conducted, where two variable-length data packets are transmitted over a 50-km dispersion-compensated span and switched at a forwarding node. The SAC labels are successfully recognized, and we obtain error-free transmission for the switched packets with less than 0.3-dB penalty.

Real-time PON signaling for emerging low-latency applications

In this paper, we provide a comprehensive discussion focused on the motivation and benefits of providing real-time (RT) signaling mechanisms in multi-channel PONs, outlining several emerging low-latency applications such as Smart Grid, machine-to-machine (M2M) communication, and cloud computing. We expand this discussion to provide a straightforward definition for signaling delay and define that for a signaling mechanism to be considered RT in PONs the majority of the signaling delay should be attributed to the one-way propagation time. We propose backwards compatible mechanisms capable of realizing RT signaling without necessarily requiring additional PHY-layer cost and complexity. Based on frequent signaling data bursts (SDBs) scheduled by the Optical Line Terminal (OLT) for selected upgraded Optical Network Units (ONUs), our proposed mechanisms enable pay-as-you-grow migration from conventional TDM PONs to highly flexible RT signaling for multi-channel PONs capable of supporting emerging low-latency applications. We provide simulation results of system performance with relevant benchmarks showing up to 15% delay reductions with respect to previously proposed OC (Optical Coding) RT architectures, which have already demonstrated significant performance increases with respect to conventional PONs. Finally, we emphasize the flexibility of the proposed mechanisms provided to future RT Dynamic Bandwidth Allocation (DBA) algorithms aimed at achieving efficient Quality-of-Service (QoS) requirements.

An algorithm for improving Sliding window Network Coding in TCP

Sliding-window Network Coding (NC) is a variation of Network Coding that is an addition to TCP/IP and improves the throughput of TCP on wireless networks. In this article, two implementations of a new algorithm are proposed in order to decrease the total transmission time, and to increase the decoding throughput throughout the transmission. The algorithms main process identifies then retransmits the number of outstanding lost packets and is implemented in two different ways. The End of Transmission (EOT) implementation applies the process only once at the end of the transmission, whereas the “Pseudo-block” (PB) implementation applies the process at regular intervals throughout file transmission. The discrete event simulator ns-2 is used to implement and test the benefits of the proposed algorithm. Our extensive simulation results show that both implementations provide a sizeable decrease in average transmission time. For the first implementation (EOT), the average time to receive data decreased by 8.04% for small files (under 1 MB) compared to TCP/NC. The second implementation, PB, reduces file transmission times by up to 70% for larger files (GB range). Furthermore, PB creates a more even decoding throughput and allows for a smoother transmission. In this work, PB is shown to decrease the average standard deviation of the decoding throughput by over 60%. This decrease in decoding delay demonstrates the potential of sliding window NC in future streaming applications.

Broadcasting XORs: On the Application of Network Coding in Access Point-to-Multipoint Networks

We investigate network coding (NC) in access point-to-multi-point (PMP) broadcast networks. Characterized by a shared unicast upstream channel and a time-shared broadcast downstream channel, PMP networks are widely deployed in optical and wireless access networks. We develop a queuing-theoretic model of NC at the medium access control (MAC) sublayer and analyze the impact of NC on packet delay. Our analysis is validated through discrete-event simulation and demonstrates significant delay advantages for NC under high loads and localized traffic.