Anantha Simha

Optimal controller placement in Software Defined Networks (SDN) using a non-zero-sum game

In this paper, we discuss optimal controller placement for Software Defined Networks (SDN) and propose a non-zero-sum game based distributed technique. Our proposed technique is a simple and low-complexity solution which runs in real-time. This can be implemented as an optimization engine at each SDN controller. The optimization engine at each controller computes a payoff function and compares its own payoff value with that of neighbors and takes appropriate decisions such that either new controllers should be added, or existing controllers should be deleted or offloading should be performed between controllers dynamically. We have conducted extensive simulations and verified the usability of the proposed scheme. We also propose a deployment framework which can be implemented using OpenFlow enabled platforms. Use of this technique not only can improve Quality of Services (QoS - minimum packet drops and delay) but also can save cost of deployment and operation.

Universal caching model and Markov-based cache analysis for information centric networks

Information Centric Networks (ICN) is proposed as an alternative design for content-based communication rather than host-to-host-based communication. In ICN architecture, the data are cached or stored at intermediate routers to satisfy future demands locally without repeatedly getting it from the original source. For ICN architecture to work efficiently, caching policies must be defined such that most of the requests are satisfied locally. In this paper, we propose Universal Caching, a novel caching algorithm which takes into account multiple parameters like distance from the content source, frequency of fetching the content, number of outgoing links at the intermediate router. We further propose a discrete time Markov model and entropy-based analysis to determine the cache size for a given cache hit requirement at a router in ICN networks. We have also established that Universal Caching policy is in fact an optimal caching policy and have calculated the bounds for the same as well. To validate the proposed caching policy and the analytical model, we have conducted extensive MATLAB-based simulations. Our simulation results show that the Universal Caching outperforms other caching policies like Least Recently Used, First In, First Out in terms of cache hits, access delay, etc.

A novel overlay architecture for Information Centric Networking

Information Centric Networks (ICN) has emerged as a new Internet paradigm where the users are being provided seamless access to contents rather than the host as in classical approach. For ICN realization, the current Internet requires architectural changes. These changes must be incorporated such that existing applications continue running without any difficulties. In this paper, we introduce a novel ICN architecture, called Overlay Information Centric Networks (O-ICN), which separates the data plane from the control plane activities of the network. Unlike the usual ICN architectures, in our approach naming and routing (control plane) functionalities have been segregated from the caching (data plane) functionality at each ICN router. We introduce a new module called an ICN Manager which is an extended version of Domain Name System (DNS) server and performs name resolution along with routing functionalities for ICN type of requests as well as the normal DNS resolution for non-ICN type of requests. The O-ICN architecture can be deployed incrementally over the current Internet and can operate both as ICN and non-ICN/legacy network simultaneously. This can also be implemented both in Software Defined Networks (SDN) as well as non-SDN networks without much modifications. The proposed architecture provides benefits in terms of simplification of network management, vendor independent solution, better use of network resources in terms of bandwidth and delay, and reduction of deployment/operational cost of the network.

Path Loss model for Indian terrain - empirical approach

It has been observed that the existing path loss models such as Hata, Okumura, and COST-231-Walfisch-Ikegami do not help in accurately predicting path loss for Indian terrain conditions. Since path loss values impacts the cost of operation along with Quality of Services (QoS) significantly, a new path loss model is desired. Modelling an accurate path loss model for Indian terrain requires terrain/city planning along with the building data, extensive drive tests for radio signal characterization, wireless telecom operator data related to Base Station (BTS) location, transmission power, height, tilt, etc. All these factors make the path loss modeling difficult and challenging. In this direction, we have proposed Tata Dense Urban Path loss (T-DUP) model which is a non-deterministic statistical model applicable for Indian cities. The proposed path loss model is validated through extensive drive tests at Bangalore with the help of a major Indian Telecom Operator operating at 1800 MHz. We have observed that the T-DUP model is statistically robust and is closely matching to the drive test results. We have also conducted statistical analysis and have illustrated the impact of various city planning parameters such as building heights, building separation, road widths, etc., on path loss model. We believe that T-DUP model can be extended to other frequency bands operating in India with appropriate tuning of parameters.

Interference-aware discovery and optimal uplink scheduling for D2D communication in LTE networks

Device-to-Device (D2D) communication underlaying a cellular infrastructure has recently gained tremendous interest not only by the researchers but also by cellular operators and equipment manufacturers. In this paper, we have introduced a novel D2D communication scheme which can be realized by a D2D discovery mechanism along with a two-phase interference-aware optimal D2D uplink scheduling scheme for Long Term Evolution (LTE) networks. Our proposed scheme is different from the D2D emergency communication use-case being proposed by 3rd Generation Partnership Project (3GPP) in (i) reuse of Resource Blocks (RBs) and (ii) network assisted D2D communication, because of which commercial use of the proposed scheme can be made possible. To validate the proposed scheme, we have conducted extensive simulations usingMatlab and emulations in a test-bed created for this purpose. Both simulation and emulation results demonstrate that by allowing D2D communication along with cellular communication, network performance is improved significantly. However, implementing D2D along with the cellular transmissions requires certain changes in the 3GPP standards and should be reflected in the future releases of LTE.

Novel Transport Layer Aware Uplink Scheduling Scheme for LTE-based Networks

In this paper we propose a novel Transport Layer Aware Uplink Scheduling Scheme (TraLAUS) for Long Term Evolution (LTE) Networks. It is a cross-layer scheduling scheme in which scheduling decisions are taken at the MAC (Medium Access Control) Layer of the TCP/IP Protocol stack based on Physical and Transport Layer informations. In this scheme, the eNodeB takes the uplink scheduling decisions based on the demand and CQI (Channel Quality Information) received from the UEs (User Equipments). Through extensive simulations, we compare the performance of our proposed algorithm with that of classical Round Robin (RR) and channel-aware Modified Round Robin (MRR) uplink scheduling schemes. From the simulation results, we observe that our algorithm performs significantly better in terms of average throughput (gain of 18.89% over RR and 13.89% over MRR), packet drops and queue length as compared to that of RR and MRR uplink scheduling schemes, and at the same time ensures fairness among the UEs.

Performance evaluation of caching policies in NDN - an ICN architecture

Information Centric Networking (ICN) advocates the philosophy of accessing the content independent of its location. Owing to this location independence in ICN, the routers en-route can be enabled to cache the content to serve the future requests for the same content locally. Several ICN architectures have been proposed in the literature along with various caching algorithms for caching and cache replacement at the routers en-route. The aim of this paper is to critically evaluate various caching policies using Named Data Networking (NDN), an ICN architecture proposed in literature. We have presented the performance comparison of different caching policies naming First In First Out (FIFO), Least Recently Used (LRU), and Universal Caching (UC) in two network models; Watts-Strogatz (WS) model (suitable for dense short link networks such as sensor networks) and Sprint topology (better suited for large Internet Service Provider (ISP) networks) using ndnSIM, an ns3 based discrete event simulator for NDN architecture. Our results indicate that UC outperforms other caching policies such as LRU and FIFO and makes UC a better alternative for both sensor networks and ISP networks.

Resource Constrained Offloading in Fog Computing

When focusing on the Internet of Things (IoT), communicating and coordinating sensor--actuator data via the cloud involves inefficient overheads and reduces autonomous behavior. The Fog Computing paradigm essentially moves the compute nodes closer to sensing entities by exploiting peers and intermediary network devices. This reduces centralized communication with the cloud and entails increased coordination between sensing entities and (possibly available) smart network gateway devices. In this paper, we analyze the utility of offloading computation among peers when working in fog based deployments. It is important to study the trade-offs involved with such computation offloading, as we deal with resource (energy, computation capacity) limited devices. Devices computing in a distributed environment may choose to locally compute part of their data and communicate the remainder to their peers. An optimization formulation is presented that is applied to various deployment scenarios, taking the computation and communication overheads into account. Our technique is demonstrated on a network of robotic sensor--actuators developed on the ROS (Robot Operating System) platform, that coordinate over the fog to complete a task. We demonstrate 77.8% latency and 54% battery usage improvements over large computation tasks, by applying this optimal offloading.

Fair and Efficient Listen Before Talk (LBT) Technique for LTE Licensed Assisted Access (LAA) Networks

Licensed Assisted Access (LAA) has been proposedby 3rd Generation Partnership Project (3GPP) in order touse Long Term Evolution (LTE) in the un-licensed band. Asthe centralized control of LAA has to co-exist with existingdistributed Wireless Fidelity (WiFi), efficient Listen Before Talk(LBT) technique is required to reduce interference and improvespectral efficiency. In this paper, we have proposed an LBTmechanism which attempts to share the medium in a fair mannerand improves the WiFi as well as combined (WiFi-LAA) systemperformances significantly. We have analyzed for enhanced Dis-tributed Coordination Function (DCF) model for LAA. We havealso observed network performance of the proposed LBT modelfor WiFi-LAA co-existence in the un-licensed spectrum usingNetwork Simulator (NS-3). The simulation based observationsare very much similar to that of the Matlab based analyticalresults. Also, observed results indicate that with proper choiceof LAA channel occupancy and back-off counter, WiFi as wellas overall system gain can be achieved.

Energy and spectral efficient direct Machine-to-Machine (M2M) communication for cellular Internet of Things (IoT) networks

Machine-to-Machine (M2M) communication is expected to be one of the major application areas of cellular communications in near future. In cellular M2M communication, although, the gateway layer provides smooth management to underlaying cellular networks, it over-burdens the network in terms of delay and bandwidth utilization. Therefore, with the recent advancements in direct Device-to-Device (D2D) communication in Long Term Evolution (LTE) networks, it becomes an effective strategy to use D2D techniques in cellular M2M networks. In this paper, we have proposed a novel direct M2M technique which uses D2D communication and offloading principles. In the proposed scheme, instead of communicating via gateway nodes, close proximity M2M devices can directly communicate among themselves using D2D communication. The proposed method shows promising gain in Quality of Services (QoS) in terms of delay and throughput for cellular M2M networks in addition to the total energy consumption. The proposed direct M2M communication technique can also be used to realize the smart technologies such as smart home/ village/city in an energy and spectral efficient way.