Ganesh Chennimala Sankaran

Performance analysis of Dynamic Bandwidth Allocation algorithms for long-reach PONs

This paper presents an analysis of existing Dynamic Bandwidth Allocation (DBA) algorithms, when applied to Long Reach Passive Optical Networks (LR-PON). LR-PONs provide coverage beyond 20 Km range up to 200 Km thus paving the way for a combined metropolitan and access network. The DBA algorithm is used to efficiently share the upstream optical bandwidth among the Optical Network Units (ONU). The objective of the analysis is to identify some critical issues in designing DBAs for LR-PONs.

Optical Traffic Grooming-Based Data Center Networks: Node Architecture and Comparison

With data center network traffic growing significantly, power-efficient optical and hybrid optical architectures are considered as an alternate to packet switching. Typically, hybrid optical architectures use fast optical switching elements to ensure any-to-any route reachability. In this paper, an optically groomed data center network (OGDCN) framework is proposed. The proposed framework supports any-to-any route reachability without using fast optical switching elements in the network. Different components can be combined to realize an OGDCN, as described in this paper. The framework is evaluated and compared to other architectures in terms of scalability and power consumption. A particular OGDCN realization is presented and is shown to be better than other architectures in terms of power consumption. The power consumption is lower by at least 47% than the next best existing architecture proposed in the literature.

ONU Buffer Elimination for Power Savings in Passive Optical Networks

In this paper, we examine the effects of a power saving scheme in Passive Optical Networks (PON). Buffers in network equipment have been a cause of concern since they consume significant power. In a PON, the Optical Network Unit (ONU) buffers client node packets before forwarding to the Optical Line Terminal (OLT). We propose a scheme that reduces or eliminates buffers in the ONU thereby providing reduction in cost and power. However, this can potentially increase the packet delay due to buffering at the client node. This paper analyzes the delay performance using simulated and theoretical models. The increase in delay due to buffering at the EN instead of the ONU is found to be less than 300 microseconds, for the studied workloads; the total packet delay is also less than 800 microseconds. This is within the access network delay budget (typically 2 milliseconds) for different traffic types and loads.

A survey of hybrid optical data center network architectures

This paper presents a survey of data center network architectures that use both optical and packet switching components. Various proposed architectures and their corresponding network operation details are discussed. Electronic processing-based packet switch architectures and hybrid optical–electronic-based switch architectures are presented. These hybrid optical switch architectures use optical switching elements in addition to traditional electronic processing entities. The choice of components used for realizing functionality including the network interfaces, buffers, lookup elements and the switching fabrics have been analyzed. These component choices are summarized for different architectures. A qualitative comparison of the various architectures is also presented.

Time synchronization mechanisms for an optically groomed data center network

This paper investigates the time synchronization aspect of transmission scheduling in an optically groomed data center network (OGDCN). The architecture is based on a hybrid optical-packet approach and uses broadcast domains and wavelength division multiplexing for communication. The salient feature of this architecture is that all network paths are readily available in the optical domain and there is no need for optical path establishment. A source-destination pair must tune to a predefined wavelength at a scheduled time during data transfer. Every compute and storage node (CSN) is equipped with one or more tunable optical transceivers. As with any broadcast network, multiple transmitters that share a link segment cannot use the same wavelength on a link at the same time since this would result in collisions. Hence, transmission scheduling is required to prevent collisions and to allot an exclusive time duration for every transmission request. This papers focus is on the time synchronization aspect that is critical for scheduling. Two schemes — continuous and discrete (slotted) time — are defined and evaluated. The objective of this paper is to understand the various factors affecting performance of these synchronization mechanisms. With the former scheme, clock accuracy has a significant impact on performance. With the latter scheme, propagation delay variance and packet length distribution impact performance in terms of utilization. The paper presents an evaluation of the performance of these mechanisms in the context of the OGDCN architecture. The results show that continuous time is able to better efficiently utilize network resources.

Scheduling in data center networks with optical traffic grooming

An optically groomed data center network (OGDCN) is a hybrid optical data center network that uses optical combiners, optical wavelength routers and packet switching components. In this architecture, each compute and storage node (CSN) is equipped with one or more tunable optical transceivers. Communications is enabled by having a sender and receiver tune to the same wavelength at the same time. Multiple transmitters using shared link segments when tuned to the same wavelength at the same time will result in collisions. Scheduling is required to prevent collisions across the network and to allot time slots to satisfy packet demands and coordination. We show that this scheduling problem is NP-complete and present two heuristics to solve this problem in a centralized server. The first heuristic (CTEF) assumes continuous time and computes the earliest free time duration for communication. The second heuristic (SLIT) uses slotted time and intersecting time within the allotted time slot for communication. The network utilization achieved by these heuristics is analyzed. It is shown that the CTEF and SLIT heuristics attain 99% and 98% utilization respectively.

Reporting in ONUs with reduced buffers

In this work, we consider a passive optical network (PON) with reduced buffer capacity at the Optical Network Unit (ONU). The objective is to reduce the ONU power consumption by having smaller buffers. This requires that some of the end-nodes packets be buffered at the end-node itself (EN) and some buffered at the ONU. These packets will be used to fill the upstream transmission time slot allotted to the ONU by the optical line terminal (OLT). Thus, the REPORT duration intimated by the ONU to the OLT should consider the EN buffer backlogs. This problem is similar to Minimum Makespan scheduling (MMS) with multiple processors. Since this problem is NP-Hard, we have considered three heuristics for computing the transit queue lengths and three sequencing strategies for filling the buffers at the ONU. With the help of simulation based performance studied, it is observed that the choice of sequencing strategy dominates the network performance followed by the choice of heuristics. The sequencing strategies that aggressively buffer packets at the ONU result in higher packet delay.

Combinatorial approach for network switch design in data center networks

This paper deals with the efficient design of network switch/routers for an optical data center network. Each switch has multiple components such as ingress/egress interfaces, optical and/or electronic buffers, interconnection switching fabric and so on. There are several possible choices available for each of these components. This paper presents a systematic approach to designing the switch architecture using a combination of these component choices, while meeting specified design criteria. It requires formally defining the structure of a switch and enforcing semantics across components. This is formulated as a constraint optimization problem with formal language grammar guiding its search process. This problem formulation is used to identify the best-possible architecture for a hierarchical DCN. Two of the three solutions identified were new and were not reported in literature. These solutions were also validated experimentally.

ONU-wavelength grouping scheme for efficient scheduling in Long Reach-PONs

Long Reach Passive Optical Networks (LR-PON) are capable of supporting upto 2048 subscribers located at a distance of 200 Km from the service provider. The higher propagation delay between the OLT and the ONU (compared to traditional PONs) leads to increased data packet delay. In case of round-robin scheduling among N ONUs, the grant for the first ONU cannot be sent until the Nth ONU has been scheduled. To mitigate this problem, a scheduling approach in which the ONUs and the wavelengths are grouped is proposed in a hybrid TDM/WDM PON environment. A number of ONUs are grouped together and are assigned a set of wavelengths dedicated for the group. The grant scheduling of ONUs in different groups is done simultaneously in parallel by the OLT. Simulation based results indicate that the proposed mechanism provides lower delay when compared to NA+ scheme consistently and performs better than Multi-thread Polling at higher loads.

Optimal Power Sharing Control in Networked Fuel Cell Stacks

Abstract Optimum use of available energy sources is essential for cost effective and sustainable growth. Fuel cells - due to their ability in efficiently extracting energy from fuels - have gained considerable attention among the various energy conversion alternatives. Systems researchers working in the field of fuel cells have been focusing on optimal stack design and the attendant modeling aspects. In a power network where multiple fuel cell stacks combine together to achieve the required power, it is not enough to focus only on the optimal design of the stacks. While operating the stacks, the problem of optimal sharing of power between the different stacks in a power network is another important problem that needs to be addressed. This optimal power sharing problem is the focus of this paper. We will describe a novel solution approach for this optimization problem, which through prior off-line computations reduces the on-line optimization task to one of solving simple equations. Another major significance of this new approach is that unlike the conventional optimizers, global optimum is guaranteed using this approach. The proposed algorithm uses a data-based model between the voltage and current for optimization. To account for changes in the system characteristics with time, a model updater algorithm that updates the data-based model using newly available data and the previous model is proposed.