检索结果-内蒙古大学图书馆

Efficient Syndrome Decoder for Heavy Hexagonal QECC via Machine Learning

学校读者我要写书评

暂无评论

arXiv 2022年

作者： Bhoumik, Debasmita Majumdar, Ritajit Madan, Dhiraj Vinayagamurthy, Dhinakaran Raghunathan, Shesha Sur-Kolay, Susmita Advanced Computing & Microelectronics Unit Indian Statistical Institute India IBM Quantum IBM India Research Lab India

Error syndromes for heavy hexagonal code and other topological codes such as surface code have typically been decoded by using Minimum Weight Perfect Matching (MWPM) based methods. Recent advances have shown that topological codes can be efficiently decoded by deploying machine learning (ML) techniques, in particular with neural networks. In this work, we first propose an ML based decoder for heavy hexagonal code and establish its efficiency in terms of the values of threshold and pseudo-threshold, for various noise models. We show that the proposed ML based decoding method achieves ∼ 5× higher values of threshold than that for MWPM. Next, exploiting the property of subsystem codes, we define gauge equivalence for heavy hexagonal code, by which two distinct errors can belong to the same error class. A linear search based method is proposed for determining the equivalent error classes. This provides a quadratic reduction in the number of error classes to be considered for both bit flip and phase flip errors, and thus a further improvement of ∼ 14% in the threshold over the basic ML decoder. Lastly, a novel technique based on rank to determine the equivalent error classes is presented, which is empirically faster than the one based on linear search. © 2022, CC BY.

关键词： Gages

A Framework for Analyzing Resource Allocation Policies for Multi-Access Edge computing

学校读者我要写书评

暂无评论

A Framework for Analyzing Resource Allocation Policies for M...

IEEE International Conference on Edge computing (EDGE)

作者： Kaustabha Ray Ansuman Banerjee Advanced Computing and Microelectronics Unit Indian Statistical Institute

ISBN: (纸本)9781665400633

Multi-Access Edge computing (MEC) is a promising new paradigm enabling low-latency access to services deployed on edge servers. This helps to avert network latencies often encountered in accessing cloud services. The cornerstone of a MEC environment is a resource allocation policy used to partition and allocate computational resources such as bandwidth, memory available on the edge server to user service invocations availing such services. In this work, we propose a generic data-driven framework to model and analyze such MEC resource allocation policies. We model a MEC system as a Turn-Based Stochastic Multi-Player Game and use Probabilistic Model Checking to derive quantitative guarantees on resource allocation policies against requirements expressed in Probabilistic Alternating-Time Temporal Logic with Rewards. We present results on state-of-the-art MEC resource allocation policies to demonstrate the effectiveness of our framework.

关键词： Analytical models Multi-access edge computing Computational modeling Stochastic processes Games Model checking Probabilistic logic

Efficient and Fault-Tolerant Memristive Neural Networks with In-Situ Training

学校读者我要写书评

暂无评论

SSRN

SSRN 2025年

作者： Prajapati, Santlal Mondal, Manobendra Nath Sur-Kolay, Susmita Indian Statistical Institute Advanced Computing and Microelectronics Unit West Bengal Kolkata700108 India UPES School of Computer Science Uttarakhand Dehradun248007 India Ashoka University Department of Computer Science Haryana Sonipat131029 India

In an artificial neural network (ANN), the most frequent and demanding operations are matrix-vector multiplication, vector-vector outer product, and weight matrix updates. These operations require parallel and in-memory processing hardware architectures to accelerate the training of the ANN. This article introduces a novel memristor-based multi-layer neural network (memristive MLNN) architecture featuring an efficient in-situ supervised training method based on gradient descent backpropagation. The proposed memristive MLNN can execute the aforementioned operations in O(1) time by leveraging in-memory and parallel computing capabilities. A single memristor without transistors is sufficient to realize a synapse. The performance of our architecture is evaluated by training and testing the memristive MLNN using the LTspice simulator, which achieved accuracies of 98.22%, 90.43%, and 98.59% on the IRIS, NASA asteroid, and Breast Cancer Wisconsin data sets respectively. The resilience of the architecture to memristor faults owing to a stuck-at-a-conducting state is assessed by injecting faults at certain percentages of randomly chosen memristors. The impact of nonlinearity, present in the memristor's conductance, on the accuracy of memristive MLNN is analyzed. We also evaluate the memristive MLNN's performance with a 10% variation in memristors. The simulation results establish that the network's performance is not affected significantly by non-linearity, device variation, and faulty memristors. © 2025, The Authors. All rights reserved.

关键词： NASA

Optimal Gathering over Weber Meeting Nodes in Infinite Grid

学校读者我要写书评

暂无评论

arXiv 2022年

作者： Bhagat, Subhash Chakraborty, Abhinav Das, Bibhuti Mukhopadhyaya, Krishnendu National Institute of Science Education and Research HBNI Bhubaneswar India Advanced Computing and Microelectronics Unit Indian Statistical Institute Kolkata India

The gathering over meeting nodes problem requires the robots to gather at one of the predefined meeting nodes. This paper investigates the problem with respect to the objective function that minimizes the total number of moves made by all the robots. In other words, the sum of the distances traveled by all the robots is minimized while accomplishing the gathering task. The robots are deployed on the nodes of an anonymous two-dimensional infinite grid which has a subset of nodes marked as meeting nodes. The robots do not agree on a global coordinate system and operate under an asynchronous scheduler. A deterministic distributed algorithm has been proposed to solve the problem for all those solvable configurations, and the initial configurations for which the problem is unsolvable have been characterized. The proposed gathering algorithm is optimal with respect to the total number of moves performed by all the robots in order to finalize the gathering. Copyright © 2022, The Authors. All rights reserved.

关键词： Robots

On the Span of l Distance Coloring of Infinite Hexagonal Grid

学校读者我要写书评

暂无评论

arXiv 2022年

作者： Ghosh, Sasthi C. Koley, Subhasis Advanced Computing and Microelectronics Unit Indian Statistical Institute 203 B. T. Road Kolkata700108 India

For a graph G(V, E) and l ∈ , an l distance coloring is a coloring f : V → {1, 2, · · ·, n} of V such that ∀u, v ∈ V, u ≠ v, f(u) ≠ f(v) when d(u, v) ≤ l. Here d(u, v) is the distance between u an... 详细信息

For a graph G(V, E) and l ∈ , an l distance coloring is a coloring f : V → {1, 2, · · ·, n} of V such that ∀u, v ∈ V, u ≠ v, f(u) ≠ f(v) when d(u, v) ≤ l. Here d(u, v) is the distance between u and v and is equal to the minimum number of edges that connect u and v in G. The span of l distance coloring of G, λl(G), is the minimum n among all l distance coloring of G. A class of channel assignment problem in cellular network can be formulated as a distance graph coloring problem in regular grid graphs. The cellular network is often modelled as an infinite hexagonal grid TH, and hence determining λl(TH) has relevance from practical point of view. Jacko and Jendrol [Discussiones Mathematicae Graph Theory, 2005] determined the exact value of λl(TH) for any odd l and for even l ≥ 8, it is conjectured that (Equation presented) where [x] is an integer, x ∈ and (Equation presented). For l = 8, the conjecture has been proved by Sasthi and Subhasis [22nd Italian Conference on Theoretical Computer Science, 2021]. In this paper, we prove the conjecture for any l ≥ 10. © 2022, CC BY.

关键词： Coloring

Optimal L(1,2)-edge Labeling of Infinite Octagonal Grid

学校读者我要写书评

暂无评论

arXiv 2022年

作者： Koley, Subhasis Ghosh, Sasthi C. Advanced Computing and Microelectronics unit Indian Statistical Institute 203 B. T. Road Kolkata700108 India

For two given non-negative integers h and k, an L(h, k)-edge labeling of a graph G = (V(G),E(G)) is a function f0 : E(G) −→ {0,1,···,n} such that ∀e1,e2 ∈ E(G), |f0(e1) − f0(e2)| ≥ h when d0(e1,e2) = 1 and |f0(e1) − f0(e2)| ≥ k when d0(e1,e2) = 2 where d0(e1,e2) denotes the distance between e1 and e2 in G. Here d0(e1,e2) = k0 if there are at least (k0 − 1) number of edges in E(G) to connect e1 and e2 in G. The objective is to find span which is the minimum n over all such L(h,k)-edge labeling and is denoted as λ0h,k(G). Motivated by the channel assignment problem in wireless cellular network, L(h,k)-edge labeling problem has been studied in various infinite regular grids. For infinite regular octagonal grid T8, it was proved that 25 ≤ λ01,2(T8) ≤ 28 [Tiziana Calamoneri, International Journal of Foundations of Computer Science, Vol. 26, No. 04, 2015] with a gap between lower and upper bounds. In this paper we fill the gap and prove that λ01,2(T8) = 28. © 2022, CC BY.

关键词： Combinatorial optimization

Stitch-avoiding Global Routing for Multiple E-Beam Lithography

学校读者我要写书评

暂无评论

Stitch-avoiding Global Routing for Multiple E-Beam Lithograp...

International Conference on VLSI Design

作者： Kritanta Saha Sudipta Paul Pritha Banerjee Susmita Sur-Kolay Advanced Computing and Microelectronics Unit Indian Statistical Institute Kolkata India Department of Computer Science and Engineering University of Calcutta Kolkata India

ISBN: (数字)9781665485050

ISBN: (纸本)9781665496612

In Multiple E-Beam Lithography (MEBL), a layout of an IC is partitioned into vertical stripes whose boundaries are called stitch-lines. Routing patterns in each stripe are written by different beams and/or in different writing passes. However, patterns cut by stitch-lines suffer from overlay errors resulting in severe pattern distortions. These distortions are mainly due to Via violation, Vertical Routing violation, or Short Polygon. In this paper, we have proposed a Stitch-avoiding Global Router. The proposed method selects the best possible global routing path for each net so that even a traditional detailed router can generate a Detailed Routing solution with fewer violations. Experimental results show that our proposed global router followed by a conventional detailed router generates solutions with 17.3%, 6.21% and 1.89% reduction in Via violations, Short polygon violations and routed wirelength respectively with no vertical routing violations, compared to those by traditional global routing.

关键词：

An RIS Deployment Strategy to Overcome Static Obstacles in Millimeter Wave D2D Communication

学校读者我要写书评

暂无评论

An RIS Deployment Strategy to Overcome Static Obstacles in M...

IEEE International Symposium on Network computing and Applications

作者： Souvik Deb Sasthi C. Ghosh Advanced Computing & Microelectronics Unit Indian Statistical Institute Kolkata India

ISBN: (纸本)9781665495516

Use of millimeter waves for short range device to device (D2D) communication is one of the most promising ways to provide high data rates. But millimeter waves are susceptible to blockages due to high penetration losses and hence require almost a line of sight (LoS) communication to provide such high data rates. When the direct LoS of a pair of devices is blocked by an obstacle, a reconfigurable intelligent surfaces (RIS) in the vicinity can be exploited to bypass the obstacle and achieve an indirect LoS between them. In this paper, we develop a strategy for placing minimum number of RISs in an environment filled with obstacles such that any pair of devices in the environment who does not have direct LoS can communicate via these RISs by establishing an indirect LoS. We convert this RIS placement problem to a set cover problem and give a greedy algorithm to get a near optimal but practical solution. This is done as a preprocessing step before the actual devices are deployed in the region. Next we present a strategy which knowing the actual deployment of the devices, selects a subset of the deployed RISs for operation together with their channel coefficients and phase shifts such that the system sum throughput of the devices is maximised. Finally simulation results show that strategically placed RISs can significantly improve system throughput in comparison to the existing deployment strategies.

关键词： Performance evaluation Greedy algorithms Millimeter wave devices Surface waves Simulation Reconfigurable intelligent surfaces Throughput

An Energy Efficient Component Carrier Selection Mechanism for LTE-NR Dual Connectivity

学校读者我要写书评

暂无评论

An Energy Efficient Component Carrier Selection Mechanism fo...

IEEE International Symposium on Network computing and Applications

作者： Shankar K. Ghosh Sasthi C. Ghosh Advanced Computing & Microelectronics Unit Indian Statistical Institute Kolkata India

ISBN: (纸本)9781665495516

In recent past, a new dual connectivity (DC) mechanism has been standardized to deal with two different radio access technologies, namely long term evolution advanced (LTE-A) and new radio (NR). The DC mechanism calls for a higher number of component carrier (CC) allocation which in turn causes increased power consumption. Moreover, to maximize the gain in DC mode, throughput received from the serving base stations need to be nearly equal. The existing CC allocation mechanisms do not take care of these facts. In this work, we first formulate the CC selection problem for LTE-NR DC as an integer linear program (ILP) which accounts load balancing across the serving BSs while minimizing CC allocation. The considered CC selection problem has also been shown to be NP hard. Due to such characteristics, we propose a suboptimal greedy algorithm to solve the CC selection problem. Simulation results confirm that our proposed algorithm outperforms the existing CC selection algorithms.

关键词： Greedy algorithms Power demand Simulation Performance gain Throughput Load management Energy efficiency