-
Energy-efficient Federated Learning with Dynamic Model Size Allocation
Authors:
M S Chaitanya Kumar,
Sai Satya Narayana J,
Yunkai Bao,
Xin Wang,
Steve Drew
Abstract:
Federated Learning (FL) presents a paradigm shift towards distributed model training across isolated data repositories or edge devices without explicit data sharing. Despite of its advantages, FL is inherently less efficient than centralized training models, leading to increased energy consumption and, consequently, higher carbon emissions. In this paper, we propose CAMA, a carbon-aware FL framewo…
▽ More
Federated Learning (FL) presents a paradigm shift towards distributed model training across isolated data repositories or edge devices without explicit data sharing. Despite of its advantages, FL is inherently less efficient than centralized training models, leading to increased energy consumption and, consequently, higher carbon emissions. In this paper, we propose CAMA, a carbon-aware FL framework, promoting the operation on renewable excess energy and spare computing capacity, aiming to minimize operational carbon emissions. CAMA introduces a dynamic model adaptation strategy which adapts the model sizes based on the availability of energy and computing resources. Ordered dropout is integratged to enable the aggregation with varying model sizes. Empirical evaluations on real-world energy and load traces demonstrate that our method achieves faster convergence and ensures equitable client participation, while scaling efficiently to handle large numbers of clients. The source code of CAMA is available at https://github.com/denoslab/CAMA.
△ Less
Submitted 23 November, 2024;
originally announced November 2024.
-
Visualizing Intelligent Tutor Interactions for Responsive Pedagogy
Authors:
Grace Guo,
Aishwarya Mudgal Sunil Kumar,
Adit Gupta,
Adam Coscia,
Chris MacLellan,
Alex Endert
Abstract:
Intelligent tutoring systems leverage AI models of expert learning and student knowledge to deliver personalized tutoring to students. While these intelligent tutors have demonstrated improved student learning outcomes, it is still unclear how teachers might integrate them into curriculum and course planning to support responsive pedagogy. In this paper, we conducted a design study with five teach…
▽ More
Intelligent tutoring systems leverage AI models of expert learning and student knowledge to deliver personalized tutoring to students. While these intelligent tutors have demonstrated improved student learning outcomes, it is still unclear how teachers might integrate them into curriculum and course planning to support responsive pedagogy. In this paper, we conducted a design study with five teachers who have deployed Apprentice Tutors, an intelligent tutoring platform, in their classes. We characterized their challenges around analyzing student interaction data from intelligent tutoring systems and built VisTA (Visualizations for Tutor Analytics), a visual analytics system that shows detailed provenance data across multiple coordinated views. We evaluated VisTA with the same five teachers, and found that the visualizations helped them better interpret intelligent tutor data, gain insights into student problem-solving provenance, and decide on necessary follow-up actions - such as providing students with further support or reviewing skills in the classroom. Finally, we discuss potential extensions of VisTA into sequence query and detection, as well as the potential for the visualizations to be useful for encouraging self-directed learning in students.
△ Less
Submitted 19 April, 2024;
originally announced April 2024.
-
On the Age of Status Updates in Unreliable Multi-Source M/G/1 Queueing Systems
Authors:
Muthukrishnan Senthil Kumar,
Aresh Dadlani,
Masoumeh Moradian,
Ahmad Khonsari,
Theodoros A. Tsiftsis
Abstract:
The timeliness of status message delivery in communications networks is subjective to time-varying wireless channel transmissions. In this paper, we investigate the age of information (AoI) of each source in a multi-source M/G/1 queueing update system with active server failures. In particular, we adopt the method of supplementary variables to derive a closed-form expression for the average AoI in…
▽ More
The timeliness of status message delivery in communications networks is subjective to time-varying wireless channel transmissions. In this paper, we investigate the age of information (AoI) of each source in a multi-source M/G/1 queueing update system with active server failures. In particular, we adopt the method of supplementary variables to derive a closed-form expression for the average AoI in terms of system parameters, where the server repair time follows a general distribution and the service time of packets generated by independent sources is a general random variable. Numerical results are provided to validate the effectiveness of the proposed packet serving policy under different parametric settings.
△ Less
Submitted 31 October, 2022; v1 submitted 2 August, 2022;
originally announced August 2022.
-
Efficacy of Asynchronous GPS Spoofing Against High Volume Consumer GNSS Receivers
Authors:
M. Surendra Kumar,
Gaurav S. Kasbekar,
Arnab Maity
Abstract:
The vulnerability of the Global Positioning System (GPS) against spoofing is known for quite some time. Also, the positioning and navigation of most semi-autonomous and autonomous drones are dependent on Global Navigation Satellite System (GNSS) signals. In prior work, simplistic or asynchronous GPS spoofing was found to be a simple, efficient, and effective cyber attack against L1 GPS or GNSS dep…
▽ More
The vulnerability of the Global Positioning System (GPS) against spoofing is known for quite some time. Also, the positioning and navigation of most semi-autonomous and autonomous drones are dependent on Global Navigation Satellite System (GNSS) signals. In prior work, simplistic or asynchronous GPS spoofing was found to be a simple, efficient, and effective cyber attack against L1 GPS or GNSS dependent commercial drones. In this paper, first we make some important observations on asynchronous GPS spoofing attacks on drones presented in prior research literature. Then, we design an asynchronous GPS spoofing attack plan. Next, we test the effectiveness of this attack against GNSS receivers (high volume consumer devices based on Android mobile phones) of different capabilities and a commercial drone (DJI Mavic 2 Pro) under various conditions. Finally, we present several novel insights based on the results of the tests.
△ Less
Submitted 18 June, 2022;
originally announced June 2022.
-
Co-evolution of Viral Processes and Structural Stability in Signed Social Networks
Authors:
Temirlan Kalimzhanov,
Amir Haji Ali Khamseh'i,
Aresh Dadlani,
Muthukrishnan Senthil Kumar,
Ahmad Khonsari
Abstract:
Prediction and control of spreading processes in social networks (SNs) are closely tied to the underlying connectivity patterns. Contrary to most existing efforts that exclusively focus on positive social user interactions, the impact of contagion processes on the temporal evolution of signed SNs (SSNs) with distinctive friendly (positive) and hostile (negative) relationships yet, remains largely…
▽ More
Prediction and control of spreading processes in social networks (SNs) are closely tied to the underlying connectivity patterns. Contrary to most existing efforts that exclusively focus on positive social user interactions, the impact of contagion processes on the temporal evolution of signed SNs (SSNs) with distinctive friendly (positive) and hostile (negative) relationships yet, remains largely unexplored. In this paper, we study the interplay between social link polarity and propagation of viral phenomena coupled with user alertness. In particular, we propose a novel energy model built on Heider's balance theory that relates the stochastic susceptible-alert-infected-susceptible epidemic dynamical model with the structural balance of SSNs to substantiate the trade-off between social tension and epidemic spread. Moreover, the role of hostile social links in the formation of disjoint friendly clusters of alerted and infected users is analyzed. Using three real-world SSN datasets, we further present a time-efficient algorithm to expedite the energy computation in our Monte-Carlo simulation method and show compelling insights on the effectiveness and rationality of user awareness and initial network settings in reaching structurally balanced local and global network energy states.
△ Less
Submitted 26 July, 2022; v1 submitted 28 July, 2021;
originally announced July 2021.
-
Overlay Secondary Spectrum Sharing with Independent Re-attempts in Cognitive Radios
Authors:
Muthukrishnan Senthil Kumar,
Aresh Dadlani,
Kiseon Kim,
Richard O. Afolabi
Abstract:
Opportunistic spectrum access (OSA) is a promising reform paradigm envisioned to address the issue of spectrum scarcity in cognitive radio networks (CRNs). While current models consider various aspects of the OSA scheme, the impact of retrial phenomenon in multi-channel CRNs has not yet been analyzed. In this work, we present a continuous-time Markov chain (CTMC) model in which the blocked/preempt…
▽ More
Opportunistic spectrum access (OSA) is a promising reform paradigm envisioned to address the issue of spectrum scarcity in cognitive radio networks (CRNs). While current models consider various aspects of the OSA scheme, the impact of retrial phenomenon in multi-channel CRNs has not yet been analyzed. In this work, we present a continuous-time Markov chain (CTMC) model in which the blocked/preempted secondary users (SUs) enter a finite retrial group (or orbit) and re-attempt independently for service in an exponentially distributed random manner. Taking into account the inherent retrial tendency of SUs, we numerically assess the performance of the proposed scheme in terms of dropping probability and throughput of SUs.
△ Less
Submitted 28 July, 2016;
originally announced July 2016.
-
Transient Analysis of a Resource-limited Recovery Policy for Epidemics: a Retrial Queueing Approach
Authors:
Aresh Dadlani,
Muthukrishnan Senthil Kumar,
Kiseon Kim,
Faryad Darabi Sahneh
Abstract:
Knowledge on the dynamics of standard epidemic models and their variants over complex networks has been well-established primarily in the stationary regime, with relatively little light shed on their transient behavior. In this paper, we analyze the transient characteristics of the classical susceptible-infected (SI) process with a recovery policy modeled as a state-dependent retrial queueing syst…
▽ More
Knowledge on the dynamics of standard epidemic models and their variants over complex networks has been well-established primarily in the stationary regime, with relatively little light shed on their transient behavior. In this paper, we analyze the transient characteristics of the classical susceptible-infected (SI) process with a recovery policy modeled as a state-dependent retrial queueing system in which arriving infected nodes, upon finding all the limited number of recovery units busy, join a virtual buffer and try persistently for service in order to regain susceptibility. In particular, we formulate the stochastic SI epidemic model with added retrial phenomenon as a finite continuous-time Markov chain (CTMC) and derive the Laplace transforms of the underlying transient state probability distributions and corresponding moments for a closed population of size $N$ driven by homogeneous and heterogeneous contacts. Our numerical results reveal the strong influence of infection heterogeneity and retrial frequency on the transient behavior of the model for various performance measures.
△ Less
Submitted 28 July, 2016;
originally announced July 2016.
-
Performance Analysis of an Unreliable $M/G/1$ Retrial Queue with Two-way Communication
Authors:
Muthukrishnan Senthil Kumar,
Aresh Dadlani,
Kiseon Kim
Abstract:
Efficient use of call center operators through technological innovations more often come at the expense of added operation management issues. In this paper, the stationary characteristics of an $M/G/1$ retrial queue is investigated where the single server, subject to active failures, primarily attends incoming calls and directs outgoing calls only when idle. The incoming calls arriving at the serv…
▽ More
Efficient use of call center operators through technological innovations more often come at the expense of added operation management issues. In this paper, the stationary characteristics of an $M/G/1$ retrial queue is investigated where the single server, subject to active failures, primarily attends incoming calls and directs outgoing calls only when idle. The incoming calls arriving at the server follow a Poisson arrival process, while outgoing calls are made in an exponentially distributed time. On finding the server unavailable (either busy or temporarily broken down), incoming calls intrinsically join the virtual orbit from which they re-attempt for service at exponentially distributed time intervals. The system stability condition along with probability generating functions for the joint queue length distribution of the number of calls in the orbit and the state of the server are derived and evaluated numerically in the context of mean system size, server availability, failure frequency and orbit waiting time.
△ Less
Submitted 26 December, 2017; v1 submitted 29 December, 2015;
originally announced December 2015.
-
Reliable Broadcast to A User Group with Limited Source Transmissions
Authors:
Xiaoli Xu,
Meenakshi Sundaram Gandhi Praveen Kumar,
Yong Liang Guan
Abstract:
In order to reduce the number of retransmissions and save power for the source node, we propose a two-phase coded scheme to achieve reliable broadcast from the source to a group of users with minimal source transmissions. In the first phase, the information packets are encoded with batched sparse (BATS) code, which are then broadcasted by the source node until the file can be cooperatively decoded…
▽ More
In order to reduce the number of retransmissions and save power for the source node, we propose a two-phase coded scheme to achieve reliable broadcast from the source to a group of users with minimal source transmissions. In the first phase, the information packets are encoded with batched sparse (BATS) code, which are then broadcasted by the source node until the file can be cooperatively decoded by the user group. In the second phase, each user broadcasts the re-encoded packets to its peers based on their respective received packets from the first phase, so that the file can be decoded by each individual user. The performance of the proposed scheme is analyzed and the rank distribution at the moment of decoding is derived, which is used as input for designing the optimal BATS code. Simulation results show that the proposed scheme can reduce the total number of retransmissions compared with the traditional single-phase broadcast with optimal erasure codes. Furthermore, since a large number of transmissions are shifted from the source node to the users, power consumptions at the source node is significantly reduced.
△ Less
Submitted 25 June, 2015;
originally announced June 2015.
-
A New Generalized Closed Form Expression for Average Bit Error Probability Over Rayleigh Fading Channel
Authors:
Sanjay Singh,
M. Sathish Kumar,
H. S. Mruthyunjaya
Abstract:
Except for a few simple digital modulation techniques, derivation of average bit error probability over fading channels is difficult and is an involved process. In this letter, curve fitting technique has been employed to express bit error probability over AWGN of any digital modulation scheme in terms of a simple Gaussian function. Using this Gaussian function, a generalized closed form expressio…
▽ More
Except for a few simple digital modulation techniques, derivation of average bit error probability over fading channels is difficult and is an involved process. In this letter, curve fitting technique has been employed to express bit error probability over AWGN of any digital modulation scheme in terms of a simple Gaussian function. Using this Gaussian function, a generalized closed form expression for computing average probability of bit error over Rayleigh fading channels has been derived. Excellent agreement has been found between error probabilities computed with our method and the rigorously calculated error probabilities of several digital modulation schemes.
△ Less
Submitted 30 September, 2012;
originally announced October 2012.
-
Performance Analysis of Error Control Coding Techniques for Peak-to-Average Power Ratio Reduction of Multicarrier Signals
Authors:
Sanjay Singh,
M. Sathish Kumar,
H. S. Mruthyunjaya
Abstract:
Increasing demands on high data rate mobile communications services will inevitably drive future broadband mobile communication systems toward achieving data transmission rates in excess of 100 Mbps. One of the promising technologies which can satisfy this demand on high data rate mobile communications services is the Orthogonal Frequency Division Multiplexing (OFDM) transmission technology which…
▽ More
Increasing demands on high data rate mobile communications services will inevitably drive future broadband mobile communication systems toward achieving data transmission rates in excess of 100 Mbps. One of the promising technologies which can satisfy this demand on high data rate mobile communications services is the Orthogonal Frequency Division Multiplexing (OFDM) transmission technology which falls under the general category of multicarrier modulation systems. OFDM is a spectrally efficient modulation technique that can achieve high speed data transmission over multipath fading channels without the need for powerful equalization techniques. However the price paid for this high spectral efficiency and less intensive equalization is low power efficiency. OFDM signals are very sensitive to non-linear effects due to the high peak-to-average power ratio (PAPR), which leads to the power inefficiency in the RF section of the transmitter. This paper analyzes the relation between aperiodic autocorrelation of OFDM symbols and PAPR. The paper also gives a comparative study of PAPR reduction performance of various channel coding techniques for the OFDM signals. For our study we have considered Hamming codes, cyclic codes, convolution codes, Golay and Reed-Muller codes. The results show that each of the channel coding technique has a different PAPR reduction performance. Coding technique with the highest value of PAPR reduction has been identified along with an illustration on PAPR reduction performances with respect to each code.
△ Less
Submitted 21 September, 2012;
originally announced September 2012.
-
Performance Metrics Analysis of Torus Embedded Hypercube Interconnection Network
Authors:
N. Gopalakrishna Kini,
M. Sathish Kumar,
H. S. Mruthyunjaya
Abstract:
Advantages of hypercube network and torus topology are used to derive an embedded architecture for product network known as torus embedded hypercube scalable interconnection network. This paper analyzes torus embedded hypercube network pertinent to parallel architecture. The network metrics are used to show how good embedded network can be designed for parallel computation. Network parameter ana…
▽ More
Advantages of hypercube network and torus topology are used to derive an embedded architecture for product network known as torus embedded hypercube scalable interconnection network. This paper analyzes torus embedded hypercube network pertinent to parallel architecture. The network metrics are used to show how good embedded network can be designed for parallel computation. Network parameter analysis and comparison of embedded network with basic networks is presented.
△ Less
Submitted 11 December, 2009;
originally announced December 2009.
