Pub Date : 2022-07-04DOI: 10.48550/arXiv.2203.14572
Xiao Cheng, S. Maghsudi
Fog computing has emerged as a potential solution to accommodate the explosive computational demand of mobile users. This potential mainly stems from the capacity of task offloading and allocation at the network edge, reducing the delay and improving the quality of service. However, optimizing the performance of a fog network is often challenging, when consider the distinct abilities and capacities of fog nodes. We study the distributed task allocation problem in such a heterogeneous fog computing network under noises. We formulate the problem as a social-concave game, where the players attempt to minimize their regret while converging to Nash equilibrium. We develop a no-regret strategy for task allocation. The strategy, namely bandit gradient ascent with momentum, is an online convex optimization algorithm with bandit feedback. Theoretical and numerical analysis show the superior performance of the proposed strategy for efficient task allocation compared to the state-of-the-art methods.
{"title":"Distributed Task Management in the Heterogeneous Fog: A Socially Concave Bandit Game","authors":"Xiao Cheng, S. Maghsudi","doi":"10.48550/arXiv.2203.14572","DOIUrl":"https://doi.org/10.48550/arXiv.2203.14572","url":null,"abstract":"Fog computing has emerged as a potential solution to accommodate the explosive computational demand of mobile users. This potential mainly stems from the capacity of task offloading and allocation at the network edge, reducing the delay and improving the quality of service. However, optimizing the performance of a fog network is often challenging, when consider the distinct abilities and capacities of fog nodes. We study the distributed task allocation problem in such a heterogeneous fog computing network under noises. We formulate the problem as a social-concave game, where the players attempt to minimize their regret while converging to Nash equilibrium. We develop a no-regret strategy for task allocation. The strategy, namely bandit gradient ascent with momentum, is an online convex optimization algorithm with bandit feedback. Theoretical and numerical analysis show the superior performance of the proposed strategy for efficient task allocation compared to the state-of-the-art methods.","PeriodicalId":423807,"journal":{"name":"2022 IEEE 23rd International Workshop on Signal Processing Advances in Wireless Communication (SPAWC)","volume":"36 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2022-07-04","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"126807094","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2022-07-04DOI: 10.1109/spawc51304.2022.9833972
Shayan Mohajer Hamidi, M. Mehrabi, A. Khandani, Deniz Gündüz
Federated learning (FL) is a promising technology which trains a machine learning model on edge devices in a distributed manner orchestrated by a parameter server (PS). To realize fast model aggregation, the uplink phase of FL could be carried out by over-the-air computation (OAC). On the one hand, engaging more devices in FL yields a model with higher prediction accuracy. On the other hand, the edge devices in OAC need to perform appropriate magnitude alignment to compensate for underlying channel coefficients. However, due to the limited power budget, this is not possible for devices experiencing deep fade. Consequently, these devices are excluded from the FL algorithm. In this paper, we propose a channel perturbation method so that no edge device is excluded due to experiencing deep fade. To this end, OAC is performed in multiple phases. In each phase, the radio frequency (RF) vicinity of PS’s antenna is intentionally perturbed by means of RF mirror structure coined in [1]. This yields independent realizations of channels between PS and devices in each phase. By using proper transmit scalars, all devices concurrently transmit their local model updates in each phase subject to a total power constraint. Then, the PS estimates the arithmetic sum of the local updates by properly combining the aggregated models obtained across all phases. The devices’ transmit scalars and PS’s de-noising factors can be efficiently found by solving a tractable optimization problem. Index Terms—Federated learning, over-the-air computation, edge machine learning, wireless communications.
{"title":"Over-the-Air Federated Learning Exploiting Channel Perturbation","authors":"Shayan Mohajer Hamidi, M. Mehrabi, A. Khandani, Deniz Gündüz","doi":"10.1109/spawc51304.2022.9833972","DOIUrl":"https://doi.org/10.1109/spawc51304.2022.9833972","url":null,"abstract":"Federated learning (FL) is a promising technology which trains a machine learning model on edge devices in a distributed manner orchestrated by a parameter server (PS). To realize fast model aggregation, the uplink phase of FL could be carried out by over-the-air computation (OAC). On the one hand, engaging more devices in FL yields a model with higher prediction accuracy. On the other hand, the edge devices in OAC need to perform appropriate magnitude alignment to compensate for underlying channel coefficients. However, due to the limited power budget, this is not possible for devices experiencing deep fade. Consequently, these devices are excluded from the FL algorithm. In this paper, we propose a channel perturbation method so that no edge device is excluded due to experiencing deep fade. To this end, OAC is performed in multiple phases. In each phase, the radio frequency (RF) vicinity of PS’s antenna is intentionally perturbed by means of RF mirror structure coined in [1]. This yields independent realizations of channels between PS and devices in each phase. By using proper transmit scalars, all devices concurrently transmit their local model updates in each phase subject to a total power constraint. Then, the PS estimates the arithmetic sum of the local updates by properly combining the aggregated models obtained across all phases. The devices’ transmit scalars and PS’s de-noising factors can be efficiently found by solving a tractable optimization problem. Index Terms—Federated learning, over-the-air computation, edge machine learning, wireless communications.","PeriodicalId":423807,"journal":{"name":"2022 IEEE 23rd International Workshop on Signal Processing Advances in Wireless Communication (SPAWC)","volume":"1 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2022-07-04","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"133946348","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2022-07-04DOI: 10.1109/spawc51304.2022.9834013
Yangchen Li, Ying Cui, Vincent K. N. Lau
This paper intends to optimize the overall implementing process of federated learning (FL) in practical edge computing systems. First, we present a general FL algorithm, namely GenQSGD+, whose parameters include the numbers of global and local iterations, mini-batch size, and step size sequence. Then, we analyze the convergence of GenQSGD+ with arbitrary algorithm parameters. Next, we optimize all the algorithm parameters of GenQSGD+ to minimize the energy cost under the constraints on the time cost, convergence error, and step size sequence. The resulting optimization problem is challenging due to its non-convexity and the presence of a dimension-varying vector variable and non-differentiable constraint functions. We transform the complicated problem into a more tractable nonconvex problem using the structural properties of the original problem and propose an iterative algorithm using general inner approximation (GIA) and complementary geometric programming (CGP) to obtain a KKT point. Finally, we numerically demonstrate remarkable gains of optimization-based GenQSGD+ over typical FL algorithms and the advancement of the proposed optimization framework for federated edge learning.
{"title":"An Optimization Framework for Federated Edge Learning","authors":"Yangchen Li, Ying Cui, Vincent K. N. Lau","doi":"10.1109/spawc51304.2022.9834013","DOIUrl":"https://doi.org/10.1109/spawc51304.2022.9834013","url":null,"abstract":"This paper intends to optimize the overall implementing process of federated learning (FL) in practical edge computing systems. First, we present a general FL algorithm, namely GenQSGD+, whose parameters include the numbers of global and local iterations, mini-batch size, and step size sequence. Then, we analyze the convergence of GenQSGD+ with arbitrary algorithm parameters. Next, we optimize all the algorithm parameters of GenQSGD+ to minimize the energy cost under the constraints on the time cost, convergence error, and step size sequence. The resulting optimization problem is challenging due to its non-convexity and the presence of a dimension-varying vector variable and non-differentiable constraint functions. We transform the complicated problem into a more tractable nonconvex problem using the structural properties of the original problem and propose an iterative algorithm using general inner approximation (GIA) and complementary geometric programming (CGP) to obtain a KKT point. Finally, we numerically demonstrate remarkable gains of optimization-based GenQSGD+ over typical FL algorithms and the advancement of the proposed optimization framework for federated edge learning.","PeriodicalId":423807,"journal":{"name":"2022 IEEE 23rd International Workshop on Signal Processing Advances in Wireless Communication (SPAWC)","volume":"67 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2022-07-04","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"130755980","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2022-07-04DOI: 10.1109/spawc51304.2022.9834007
I. Bergel, S. Mohajer
We present novel schemes for cache-aided communication over networks with a multi-antenna base station (BS) that serves multiple single-antenna users. The schemes are based on a greedy scheduling [1], which simultaneously transmits coded messages to disjoint groups of users. The proposed algorithms use the channel state information to opportunistically choose the groups to be served together and to allocate power to each coded message in order to minimize the overall communication delay. Numerical study shows that the new schemes outperform the previously known schemes.
{"title":"Channel Aware Greedy Algorithm for MISO Cache-Aided Communication","authors":"I. Bergel, S. Mohajer","doi":"10.1109/spawc51304.2022.9834007","DOIUrl":"https://doi.org/10.1109/spawc51304.2022.9834007","url":null,"abstract":"We present novel schemes for cache-aided communication over networks with a multi-antenna base station (BS) that serves multiple single-antenna users. The schemes are based on a greedy scheduling [1], which simultaneously transmits coded messages to disjoint groups of users. The proposed algorithms use the channel state information to opportunistically choose the groups to be served together and to allocate power to each coded message in order to minimize the overall communication delay. Numerical study shows that the new schemes outperform the previously known schemes.","PeriodicalId":423807,"journal":{"name":"2022 IEEE 23rd International Workshop on Signal Processing Advances in Wireless Communication (SPAWC)","volume":"80 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2022-07-04","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"132110606","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2022-07-04DOI: 10.1109/spawc51304.2022.9834019
T. Haustein, Jasmina Mcmenamy, L. Thiele, P.S.H. Leather
The successful deployment of Reconfigurable Intelligent Surface (RIS) technology in cellular networks is highly dependent on the availability of existing frameworks to enable the integration and operation of RIS panels in real-world propagation environments, preferably with low standardisation impact. This paper identifies the suitability of the existing 5G-NR frameworks for channel estimation, link control and beam management. The authors address the challenges related to the control of RIS infrastructure in radio environments in which multiple cellular networks have been deployed. The impact of adjacent bands and RIS operation that includes inter-cell links of another network is also discussed. The wideband reflective properties of the RIS are considered non-separable amongst operators.
{"title":"Reconfigurable Intelligent Surface Deployment in 5G and Beyond 5G Cellular Networks","authors":"T. Haustein, Jasmina Mcmenamy, L. Thiele, P.S.H. Leather","doi":"10.1109/spawc51304.2022.9834019","DOIUrl":"https://doi.org/10.1109/spawc51304.2022.9834019","url":null,"abstract":"The successful deployment of Reconfigurable Intelligent Surface (RIS) technology in cellular networks is highly dependent on the availability of existing frameworks to enable the integration and operation of RIS panels in real-world propagation environments, preferably with low standardisation impact. This paper identifies the suitability of the existing 5G-NR frameworks for channel estimation, link control and beam management. The authors address the challenges related to the control of RIS infrastructure in radio environments in which multiple cellular networks have been deployed. The impact of adjacent bands and RIS operation that includes inter-cell links of another network is also discussed. The wideband reflective properties of the RIS are considered non-separable amongst operators.","PeriodicalId":423807,"journal":{"name":"2022 IEEE 23rd International Workshop on Signal Processing Advances in Wireless Communication (SPAWC)","volume":"47 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2022-07-04","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"122196851","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2022-07-04DOI: 10.1109/spawc51304.2022.9833936
Pradosh Kumar Hota, Deepak Mishra, R. Saini, Ankit Dubey
Designing a secure non-orthogonal multiple access (NOMA) system has recently sparked a lot of interest among researchers due to its competency over bandwidth utilization. This paper addresses the critical security issue that arises from the adoption of successive interference cancellation based decoding approach in a two-user untrusted NOMA scenario. With the objective of maximizing secrecy fairness between users, a joint optimization problem of power allocation and decoding order is investigated to maximize the minimum ergodic secrecy rate between users. The optimized solutions are obtained by proposing a low computational complexity iterative method. In addition, closed-form optimal solutions based on a tight approximation are presented. Finally, numerical simulations validate the key nontrivial analytical claims and demonstrate that substantial performance gain is achieved over existing benchmark schemes.
{"title":"Secure NOMA for Maximizing Ergodic Secrecy Fairness in the Presence of Untrusted Users","authors":"Pradosh Kumar Hota, Deepak Mishra, R. Saini, Ankit Dubey","doi":"10.1109/spawc51304.2022.9833936","DOIUrl":"https://doi.org/10.1109/spawc51304.2022.9833936","url":null,"abstract":"Designing a secure non-orthogonal multiple access (NOMA) system has recently sparked a lot of interest among researchers due to its competency over bandwidth utilization. This paper addresses the critical security issue that arises from the adoption of successive interference cancellation based decoding approach in a two-user untrusted NOMA scenario. With the objective of maximizing secrecy fairness between users, a joint optimization problem of power allocation and decoding order is investigated to maximize the minimum ergodic secrecy rate between users. The optimized solutions are obtained by proposing a low computational complexity iterative method. In addition, closed-form optimal solutions based on a tight approximation are presented. Finally, numerical simulations validate the key nontrivial analytical claims and demonstrate that substantial performance gain is achieved over existing benchmark schemes.","PeriodicalId":423807,"journal":{"name":"2022 IEEE 23rd International Workshop on Signal Processing Advances in Wireless Communication (SPAWC)","volume":"72 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2022-07-04","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"127134667","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2022-07-04DOI: 10.1109/spawc51304.2022.9833953
Tobias Monsees, D. Wübben, A. Dekorsy, Oliver Griebel, M. Herrmann, N. Wehn
In this paper, we present a new design of Finite Alphabet (FA) Message Passing (MP) decoders using only integer operations. We utilize Discrete Density Evolution with a multidimensional Lookup-Table (mLUT) design for Variable Node (VN) updates to consider all input messages jointly for reducing the information loss compared to the frequent sequential LUT design approaches. From this mLUT design, we derive a minimum-integer computation (MIC) decoder that allows for different bit-widths for node operations and message exchanges between nodes. The mLUT operations for VN updates are replaced by low complexity signed integer additions and threshold operations, and the Check Node (CN) updates simplify to a minimum search over integers. For a (816,406) regular LDPC code, we show that our 3-bit MIC decoder achieves the communication performance of the corresponding mLUT decoder and outperforms a 4-bit state-of-the-art Min-Sum (MS) decoder. We show that the node implementations on a 22 nm FD-SOI technology yield an improved area and energy efficiency over the respective MS implementation. To the best of our knowledge, this is the first time that an implementation improvement for the VNs and CNs is shown when using FA MP.
{"title":"Finite-Alphabet Message Passing using only Integer Operations for Highly Parallel LDPC Decoders","authors":"Tobias Monsees, D. Wübben, A. Dekorsy, Oliver Griebel, M. Herrmann, N. Wehn","doi":"10.1109/spawc51304.2022.9833953","DOIUrl":"https://doi.org/10.1109/spawc51304.2022.9833953","url":null,"abstract":"In this paper, we present a new design of Finite Alphabet (FA) Message Passing (MP) decoders using only integer operations. We utilize Discrete Density Evolution with a multidimensional Lookup-Table (mLUT) design for Variable Node (VN) updates to consider all input messages jointly for reducing the information loss compared to the frequent sequential LUT design approaches. From this mLUT design, we derive a minimum-integer computation (MIC) decoder that allows for different bit-widths for node operations and message exchanges between nodes. The mLUT operations for VN updates are replaced by low complexity signed integer additions and threshold operations, and the Check Node (CN) updates simplify to a minimum search over integers. For a (816,406) regular LDPC code, we show that our 3-bit MIC decoder achieves the communication performance of the corresponding mLUT decoder and outperforms a 4-bit state-of-the-art Min-Sum (MS) decoder. We show that the node implementations on a 22 nm FD-SOI technology yield an improved area and energy efficiency over the respective MS implementation. To the best of our knowledge, this is the first time that an implementation improvement for the VNs and CNs is shown when using FA MP.","PeriodicalId":423807,"journal":{"name":"2022 IEEE 23rd International Workshop on Signal Processing Advances in Wireless Communication (SPAWC)","volume":"5 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2022-07-04","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"126439682","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2022-07-04DOI: 10.1109/spawc51304.2022.9834005
Stefan Baumgartner, O. Lang, M. Huemer
Model-based estimation methods have been employed for the task of equalization since the beginning of digital communications. Due to the incredible success of data-driven machine learning approaches for many applications in different research disciplines, the replacement of model-based equalization methods by neural networks has been investigated recently. Incorporating model knowledge into a neural network is a possible approach for complexity reduction and performance enhancement, which is, however, very challenging. In this paper, we propose a novel neural network architecture for single carrier systems with frequency domain equalization inspired by a model-based soft interference cancellation scheme. We evaluate its bit error ratio performance in indoor frequency selective-environments and show that the proposed approach outperforms both model-based and data-driven state-of-the-art methods.
{"title":"A Soft Interference Cancellation Inspired Neural Network for SC-FDE","authors":"Stefan Baumgartner, O. Lang, M. Huemer","doi":"10.1109/spawc51304.2022.9834005","DOIUrl":"https://doi.org/10.1109/spawc51304.2022.9834005","url":null,"abstract":"Model-based estimation methods have been employed for the task of equalization since the beginning of digital communications. Due to the incredible success of data-driven machine learning approaches for many applications in different research disciplines, the replacement of model-based equalization methods by neural networks has been investigated recently. Incorporating model knowledge into a neural network is a possible approach for complexity reduction and performance enhancement, which is, however, very challenging. In this paper, we propose a novel neural network architecture for single carrier systems with frequency domain equalization inspired by a model-based soft interference cancellation scheme. We evaluate its bit error ratio performance in indoor frequency selective-environments and show that the proposed approach outperforms both model-based and data-driven state-of-the-art methods.","PeriodicalId":423807,"journal":{"name":"2022 IEEE 23rd International Workshop on Signal Processing Advances in Wireless Communication (SPAWC)","volume":"12 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2022-07-04","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"122780184","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2022-07-04DOI: 10.1109/spawc51304.2022.9833975
Arjun Singh, Ali J. Alqaraghuli, J. Jornet
In this paper, the performance criteria, required design, and possible operating modes for IRS at THz frequencies are derived and presented. Due to device constraints, codebooks that can define wavefronts within the near-field may be required for optimal IRS efficiency. Numerical results are provided to benchmark the performance of Bessel beams with conventional beamforming under various communication scenarios, which show that Bessel beam wavefronts have promising applications in next-generation wireless standards at (sub) THz frequencies.
{"title":"Wavefront Engineering at Terahertz Frequencies Through Intelligent Reflecting Surfaces","authors":"Arjun Singh, Ali J. Alqaraghuli, J. Jornet","doi":"10.1109/spawc51304.2022.9833975","DOIUrl":"https://doi.org/10.1109/spawc51304.2022.9833975","url":null,"abstract":"In this paper, the performance criteria, required design, and possible operating modes for IRS at THz frequencies are derived and presented. Due to device constraints, codebooks that can define wavefronts within the near-field may be required for optimal IRS efficiency. Numerical results are provided to benchmark the performance of Bessel beams with conventional beamforming under various communication scenarios, which show that Bessel beam wavefronts have promising applications in next-generation wireless standards at (sub) THz frequencies.","PeriodicalId":423807,"journal":{"name":"2022 IEEE 23rd International Workshop on Signal Processing Advances in Wireless Communication (SPAWC)","volume":"17 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2022-07-04","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"121549872","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2022-07-04DOI: 10.1109/spawc51304.2022.9833974
R. Jaiswal, M. Elnourani, Siddharth Deshmukh, B. Beferull-Lozano
This paper investigates the problem of transfer learning in radio map estimation for indoor wireless communications, which can be exploited for different applications, such as channel modelling, resource allocation, network planning, and reducing the number of necessary power measurements. Due to the nature of wireless communications, a radio map model developed under a particular environment can not be directly used in a new environment because of the changes in the propagation characteristics, thus creating a new model for every environment requires in general a large amount of data and is computationally demanding. To address these issues, we design an effective novel data-driven transfer learning procedure that transfers and fine-tunes a deep neural network (DNN)-based model for a radio map learned from an original indoor wireless environment to other different indoor wireless environments. Our method allows to predict the amount of training data needed in new indoor wireless environments when performing the operation of transfer learning using our similarity measure. Our simulation results illustrate that the proposed method achieves a saving of 60-70% in sensor measurement data and is able to adapt to a new wireless environment with a small amount of additional data.
{"title":"Deep Transfer Learning Based Radio Map Estimation for Indoor Wireless Communications","authors":"R. Jaiswal, M. Elnourani, Siddharth Deshmukh, B. Beferull-Lozano","doi":"10.1109/spawc51304.2022.9833974","DOIUrl":"https://doi.org/10.1109/spawc51304.2022.9833974","url":null,"abstract":"This paper investigates the problem of transfer learning in radio map estimation for indoor wireless communications, which can be exploited for different applications, such as channel modelling, resource allocation, network planning, and reducing the number of necessary power measurements. Due to the nature of wireless communications, a radio map model developed under a particular environment can not be directly used in a new environment because of the changes in the propagation characteristics, thus creating a new model for every environment requires in general a large amount of data and is computationally demanding. To address these issues, we design an effective novel data-driven transfer learning procedure that transfers and fine-tunes a deep neural network (DNN)-based model for a radio map learned from an original indoor wireless environment to other different indoor wireless environments. Our method allows to predict the amount of training data needed in new indoor wireless environments when performing the operation of transfer learning using our similarity measure. Our simulation results illustrate that the proposed method achieves a saving of 60-70% in sensor measurement data and is able to adapt to a new wireless environment with a small amount of additional data.","PeriodicalId":423807,"journal":{"name":"2022 IEEE 23rd International Workshop on Signal Processing Advances in Wireless Communication (SPAWC)","volume":"4 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2022-07-04","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"114216801","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
京公网安备 11010802042870号
京ICP备2023020795号-1
扫码关注我们
求助内容：
应助结果提醒方式：