Shanti Chilukuri

Verified @gmail.com

Professor, Dept. of Computer Science and Engineering
GITAM



        

https://researchid.co/chilukurishanti
3

Scopus Publications

Scopus Publications

  • SmartHART: A Priority-aware Scheduling and Routing Scheme for IIoT Networks using Deep Reinforcement Learning
    Shanti Chilukuri, Aditya Gupta, and Hemanth Sri Sai Pulamolu
    IEEE
    The aims of the Industrial Internet of Things (IIoT) are improved efficiency of production facilities and operations in factories and better logistics and supply chain management. Communication of data in the IIoT can be challenging due to diverse and critical demands on the Quality of Service (QoS). Medium Access Control (MAC) in such networks is typically using Time Division Multiple Access as recommended by stan-dards such as the WirelessHART. In this paper, we propose and evaluate a Deep Reinforcement Learning (DRL)-based scheduling and routing scheme (called SmartHART) for WirelessHART networks. SmartHART leverages on redundancy in the network and chooses schedules and routes that minimize the end-to-end delay of data, taking priority of data into consideration. The reward design of the SmartHART agent also minimizes the maximum packet queue length in a network, making it suitable for memory-constrained field devices. Simulation results show that SmartHART can reduce the end-to-end delay by as much as 18% and the maximum packet queue size by up to 60%.

  • Deadline-Aware TDMA Scheduling for Multihop Networks Using Reinforcement Learning
    Shanti Chilukuri, Guangyuan Piao, Diego Lugones, and Dirk Pesch
    IEEE
    Time division multiple access (TDMA) is the medium access control strategy of choice for multihop networks with deterministic delay guarantee requirements. As such, many Internet of Things applications use protocols based on time division multiple access. Optimal slot assignment in such networks is NP-hard when there are strict deadline requirements and is generally done using heuristics that give suboptimal transmission schedules in linear time. However, existing heuristics make a scheduling decision at each time slot based on the same criterion without considering its effect on subsequent network states or scheduling actions. Here, we first identify a set of node features that capture the information necessary for network state representation to aid building schedules using Reinforcement Learning (RL). We then propose three different centralized approaches to RL-based TDMA scheduling that vary in training and network representation methods. Using RL allows applying diverse criteria at different time slots while considering the effect of a scheduling action on meeting the scheduling objective for the entire TDMA frame, resulting in better schedules. We compare the three proposed schemes in terms of how well they meet the scheduling objectives and their applicability to networks with memory and time constraints. One of the schemes proposed is RLSchedule, which is particularly suited to constrained networks. Simulation results for a variety of network scenarios show that RLSchedule reduces the percentage of packets missing deadlines by up to 60% compared to the best available baseline heuristic.

  • RECCE: Deep Reinforcement Learning for Joint Routing and Scheduling in Time-Constrained Wireless Networks
    Shanti Chilukuri and Dirk Pesch
    Institute of Electrical and Electronics Engineers (IEEE)
    Time Division Multiple Access-based Medium Access Control protocols tend to be the choice for wireless networks that require deterministic delay guarantees, as is the case in many Industrial Internet of Things (IIoT) applications. As the optimal joint scheduling and routing problem for multi-hop wireless networks is NP-hard, heuristics are generally used for building schedules. However, heuristics normally result in sub-optimal schedules, which may result in packets missing their deadlines. In this paper, we present RECCE, a deep REinforcement learning method for joint routing and sCheduling in time-ConstrainEd networks with centralised control. During training, RECCE considers multiple routes and criteria for scheduling in any given time slot and channel in a multi-channel, multi-hop wireless network. This allows RECCE to explore and learn routes and schedules to deliver more packets within the deadline. Simulation results show that RECCE can reduce the number of packets missing the deadline by as much as 55% and increase schedulability by up to 30%, both relative to the best baseline heuristic. RECCE can deal well with dynamic network conditions, performing better than the best baseline heuristic in up to 74% of the scenarios in the training set and in up to 64% of scenarios not in the training set.

  • NimbleCache - Low cost, dynamic cache allocation in constrained edge environments
    Shanti Chilukuri and Dirk Pesch
    IEEE
    Edge computing and caching of data in the Internet of Things (IoT) has several benefits such as reduced energy consumption by IoT end devices and increased availability of data and Quality of Service (QoS). In typical IoT scenarios, edge nodes (gateways) support several end devices, each of which may produce data in different patterns. In addition, data generated by different types of end devices varies in the application QoS requirements while also widely varying in the data access patterns by IoT services. Managing the data storage resources at edge nodes in such scenarios is a difficult task, especially since the edge nodes themselves may have limited computation capability and storage space. In this paper, we propose a dynamic, differentiated edge cache allocation strategy called NimbleCache that has low computational requirements and performs efficient cache allocation at edge nodes. Based on a Mixture Density Network (MDN), NimbleCache allocates varying portions of the edge cache to traffic of different IoT applications to achieve cache hit ratios very close to the target hit ratio. Simulation results show that NimbleCache achieves good average cache hit ratio with low cache space requirement and small computational overhead.

  • On the convergence and optimality of the firefly algorithm for opportunistic spectrum access
    Lakshmana Rao Kalabarige, Sireesha Rodda, and Shanti Chilukuri
    Inderscience Publishers

  • Achieving Optimal Cache Utility in Constrained Wireless Networks through Federated Learning
    Shanti Chilukuri and Dirk Pesch
    IEEE
    Edge computing allows constrained end devices in wireless networks to offioad heavy computing tasks or data storage when local resources are insufficient. Edge nodes can provide resources such as the bandwidth, storage and innetwork compute power. For example, edge nodes can provide data caches to which constrained end devices can off-load their data and from where user can access data more effectively. However, fair allocation of these resources to competing end devices and data classes while providing good Quality of Service is a challenging task, due to frequently changing network topology and/or traffic conditions. In this paper, we present Federated learning-based dynamic Cache allocation (FedCache) for edge caches in dynamic, constrained networks. FedCache uses federated learning to learn the benefit of a particular cache allocation with low communication overhead. Edge nodes learn locally to adapt to different network conditions and collaboratively share this knowledge so as to avoid having to transmit all data to a single location. Through this federated learning approach, nodes can find resource allocations that result in maximum fairness or efficiency in terms of the cache hit ratio for a given network state. Simulation results show that cache resource allocation using FedCache results in optimal fairness or efficiency of utility for different classes of data when compared to proportional allocation, while incurring low communication overhead.

  • Adaptive Differentiated Edge Caching with Machine Learning for V2X Communication
    Vinayaka Shashank Varanasi and Shanti Chilukuri
    IEEE
    Connected vehicles that communicate with the traffic network around them have several uses in providing road safety and infotainment. Such applications leverage on Vehicle-to-Anything (V2X) communication, which is challenging because of rapidly changing topology and traffic patterns. We propose a differentiated edge caching scheme called FlexiCache for such networks. In FlexiCache, the cache is split into sections to hold data of different classes with suitable replacement policies. Further, FlexiCache uses kernel ridge regression (KRR) to predict the proportion of cache to be allocated to each traffic type, for a desired quality of service(QoS) parameter. It then uses a self-learning mechanism that adapts cache allocation to the network conditions. Simulation results show that FlexiCache performs better than undifferentiated caching and also that the predictions by KRR result in QoS which is very close to the target value.

  • T-move: A light-weight protocol for improved QoS in content-centric networks with producer mobility
    Swaroopa Korla and Shanti Chilukuri
    MDPI AG
    Recent interest in applications where content is of primary interest has triggered the exploration of a variety of protocols and algorithms. For such networks that are information-centric, architectures such as the Content-Centric Networking have been proven to result in good network performance. However, such architectures are still evolving to cater for application-specific requirements. This paper proposes T-Move, a light-weight solution for producer mobility and caching at the edge that is especially suitable for content-centric networks with mobile content producers. T-Move introduces a novel concept called trendiness of data for Content-Centric Networking (CCN)/Named Data Networking (NDN)-based networks. It enhances network performance and quality of service (QoS) using two strategies—cache replacement and proactive content-pushing for handling producer mobility—both based on trendiness. It uses simple operations and smaller control message overhead and is suitable for networks where the response needs to be quick. Simulation results using ndnSIM show reduced traffic, content retrieval time, and increased cache hit ratio with T-Move, when compared to MAP-Me and plain NDN for networks of different sizes and mobility rates.

  • Supporting QoS Differentiation in Energy-Constrained Cognitive Radio Networks
    Lakshmana Rao Kalabarige and Shanti Chilukuri
    Springer Science and Business Media LLC
    Routing protocols for cognitive radio ad hoc networks (CRNs) select a route between the source and destination nodes based on the spectrum opportunity at intermediate nodes. When multiple routes are possible, most routing protocols for CRNs use some metric—independent of the traffic class—to select routes. However, a route that works well for transferring a particular class of data may not be the right one for a different data class, as its quality of service (QoS) requirements may differ. In this paper, we propose a reactive energy efficient routing protocol with differentiated services (REEDS) for cognitive radio networks. Route selection in REEDS is based on different (multiple) hop metrics calculated dynamically for different traffic classes so that a minimum level of QoS is guaranteed. Another characteristic feature of REEDS is the prediction and dodging of nodes that may be excessively loaded with traffic. This results in the avoidance of formation of holes due to heavy energy expenditure by some nodes. Simulation shows that routes in REEDS are established so that the QoS requirements of each traffic class are satisfied and lesser energy is consumed compared to other routing protocols for CRNs.

  • RainCloud - Cloudlet selection for effective cyber foraging
    Shanti Chilukuri, Sourabh Bollapragada, Sainath Kommineni, and Kalyana Chakravarthy C.
    IEEE
    Cyber foraging is a technique in which devices (like mobile phones) that have lesser computing resources make use of those at other resource-rich devices. Offloading computation to a nearby device (a cloudlet) rather than some remote server, has proved to be beneficial for several resource-intensive mobile applications (apps) like online-gaming, video streaming and face recognition. While techniques to offload applications such that the load on the cloudlets is balanced have been proposed, we believe that one rule does not fit all types of applications. In this paper, we propose RainCloud, a cloudlet selection heuristic with dynamic VM provisioning for effective cyber foraging. RainCloud ranks cloudlets based on their available system#x002F;network resources (capabilities), specifically depending on the resources required by the application. It also looks for a relatively stable cloudlet so that the app may run to completion before the cloudlet moves out of range of the mobile device. We tested this heuristic on a small testbed and verified that it improves application response time and QoS (in terms of jitter). We also noted that RainCloud reduces the startup time, which is the time taken for the transmission and synthesis of the VM required to run the app on the cloudlet.

  • Provider mobility in content centric networks – issues and challenges
    Shanti Chilukuri, Kalyana Chakravarthy Chilukuri, and Swaroopa Korla
    Inderscience Publishers

  • Delay-aware TDMA scheduling for multi-hop wireless networks
    Shanti Chilukuri and Anirudha Sahoo
    ACM
    Time Division Multiple Access (TDMA)-based medium access control (MAC) protocols can be used to provide guaranteed quality of service (QoS). Since nodes follow a fixed schedule to transmit data, the schedule plays a major role in determining QoS in terms of delay, throughput etc. In this study, we focus on end-to-end delay and bandwidth utilization as the QoS parameters. We present a TDMA scheduling scheme that minimizes the end-to-end delay of data, while reusing time slots where possible. This is done by ordering the transmissions along a path such that the scheduling delay at intermediate nodes is minimized. The scheme results in end-to-end delay spanning multiple frames. However, because the TDMA frame length is minimized, the end-to-end delay is minimized while increasing the throughput. The proposed scheme uses a heuristic which has a complexity of O(n2), making it suitable for networks with a large number of nodes. We also present a simple distributed version of our heuristic. Simulation results show that our scheme performs better than previous work in literature in terms of end-to-end delay and slot reuse.

  • The cloudlet with a silver lining
    Sainath Kommineni, Aneesh De, Sashank Alladi, and Shanti Chilukuri
    IEEE
    Cyber-foraging makes resource-constrained mobile devices more powerful by using the resources elsewhere in the network. Just-in-time cyber foraging with dynamic VM synthesis has been proposed to increase the performance of mobile devices, which connect a nearby fixed device - called a cloudlet - and borrow its capabilities. The cloudlet runs an instance of the VM of mobile device and the application (app) on it. However, off-loading itself may consume time and memory, leading to poor performance of the app. When there are many cloudlets within range of the mobile device, choosing just any cloudlet to connect to may lead to avoidable and noticeable delays and poor performance of the application. In this paper, we propose a scheme where the best possible cloudlet can be identified by the mobile device to connect to. The choice of the cloudlet should be based on several factors like its processing power, available memory and bandwidth and the nature of the app itself. By choosing a cloudlet more suited for the application as the site of cyber-foraging, we aim to improve the performance of the mobile application.

  • Energy Efficient and Reliable Transmission of Data in Wireless Sensor Networks
    Chilukuri Shanti and Anirudha Sahoo
    Springer International Publishing
    Reliable transmission of data over the lossy wireless medium can be achieved by using an acknowledgement (ack) scheme. However, acks can be quite energy-consuming and should be used sparingly in energy-constrained networks. The simplest of acknowledgement schemes is the stop-and-wait ARQ (swack). Recent work proposes an implicit acknowledgement scheme (iack) which seems to be energy-saving, as there are no separate ackowledgements sent. IEEE 802.11e proposes a hop-by-hop block ack (eack). We modify this scheme to suit WSNs and propose that the choice of the ack scheme that would keep the network in- tact for the longest time depends on the tree overlay and the data and ack packet lengths. We theoretically compare the maximum energy spent for sending and re- ceiving acknowledgements for these three types of ack schemes. We then give a guideline to choose the most energy-saving of these three types of acknowledge- ments for convergecast in a wireless sensor network (WSN) applications with a routing tree overlay and continuous event generation rate.

  • TREEFP: A TDMA-based reliable and energy efficient flooding protocol for WSNs
    Chilukuri Shanti and Anirudha Sahoo
    IEEE
    Flooding a network with a message from the sink is required for many purposes like synchronization, code dissemination etc. While several flooding schemes exist, only a few are designed to achieve the energy efficiency required by Wireless Sensor Networks (WSNs). In this paper, we present a TDMA-based Reliable and Energy Efficient Flooding Protocol (TREEFP) for WSNs. Slot assignment in TREEFP is done such that the time taken to flood the network is bounded to a single TDMA frame. TREEFP has a tunable system parameter which brings in tradeoff between reliability, flooding delay and energy consumption because when this parameter changes, the topology of the logical flooding tree also changes. We provide details of simulation experiments to compare TREEFP with other flooding protocols in the literature like FTSP, TDFS and MST. Simulation results show that TREEFP is better than FTSP and TDFS in terms of energy and flooding delay and comparable to MST in terms of those metrics. In terms of reliability, TREEFP is better than MST and comparable to FTSP.

  • Distributed fault tolerance for WSNs with routing tree overlays
    C Shanti and A Sahoo
    IEEE
    WSNs are inherently power constrained and are often deployed in harsh environments. As such, node death is a possibility that must be considered while designing protocols for such networks. Rerouting of data is generally necessary so that data from the descendant nodes of the dead node can reach the sink. Since slot allocation in TDMA MAC protocols is generally done based on the routing tree, all the nodes must switch to the new routing tree to avoid collisions. This necessitates disseminating the fault information to all the nodes reliably. We propose a flooding algorithm for disseminating fault info to the network reliably even in a lossy channel. Simulation results show that the proposed flooding scheme consumes lesser energy and converges faster than a simple flooding scheme.

  • DGRAM: A delay guaranteed routing and MAC protocol for wireless sensor networks
    Anirudha Sahoo and Shanti Chilukuri
    Institute of Electrical and Electronics Engineers (IEEE)
    This paper presents an integrated MAC and routing protocol called Delay Guaranteed Routing and MAC (DGRAM) for delay sensitive wireless sensor network (WSN) applications. DGRAM is a TDMA-based protocol designed to provide deterministic delay guarantee in an energy efficient manner. The design is based on slot reuse to reduce the latency of a node in accessing the medium, while ensuring contention free medium access. The transmission and reception cycles of nodes are carefully computed so that data is transported from the source towards the sink while the nodes could sleep at the other times to conserve energy. Thus, routes of data packets are integrated into DGRAM. We provide a detailed design of time slot assignment and delay analysis of the protocol. One major advantage of DGRAM over other TDMA protocols is that the slot assignment is done in a fully distributed manner making the DGRAM network self-configuring. We have simulated DGRAM using ns2 simulator and compared the results with those of SMAC for a similar network. Simulation results show that the delay experienced by data packets is always less than the analytical delay bound for which the protocol is designed. As per simulation results, the average energy consumption does not change as the event rate changes, and is less than that of SMAC. This characteristic of DGRAM provides flexibility in choosing various operating parameters without having to worry about energy efficiency.

  • DGRAM: A delay guaranteed routing and MAC protocol for wireless sensor networks
    Chilukuri Shanti and Anirudha Sahoo
    IEEE
    This paper presents an integrated MAC and routing protocol called Delay Guaranteed Routing and MAC (DGRAM) for delay-sensitive wireless sensor network (WSN) applications. DGRAM is a TDMA-based protocol designed to provide deterministic delay guarantee in an energy-efficient manner. The design is based on slot reuse to reduce latency of a node in accessing the medium, while ensuring that the medium access is contention-free. The transmission and reception slots of nodes are carefully computed so that data is transported from the source toward the sink while the nodes could sleep at the other times to conserve energy. Thus, routes of data packets are integrated into DGRAM, i.e., there is no need for a separate routing protocol in a DGRAM network. We provide a detailed design of time slot assignment and delay analysis of the protocol. We have simulated DGRAM using ns2 simulator and compared the results with those of FlexiTP, which is another TDMA protocol that claims to provide delay guarantee, and with those of a basic TDMA MAC. Simulation results show that the delay experienced by data packets is always less than the analytical delay bound for which the protocol is designed. Also, the TDMA frame size with DGRAM is always lesser compared to that of FlexiTP, which makes the maximum possible delay much lesser than that of FlexiTP. The average delay experienced by packets and the average total energy spent in the network are much lesser in a network using DGRAM than that using FlexiTP or the basic TDMA MAC.