The 4th Age of Information Workshop (AoI 2021)

Session AoI-S1

Session 1: AoI in Sensing and Control

9:00 AM — 10:30 AM EDT
May 10 Mon, 9:00 AM — 10:30 AM EDT

Intermittent Status Updating Through Joint Scheduling of Sensing and Retransmissions

Omur Ozel (George Washington University, USA); Parisa Rafiee (George Washington University, USA)

Consider an energy harvesting node where generation of a status update message takes non-negligible time due to sensing, computing and analytics operations performed before making update transmissions. The node has to harmonize its (re)transmission strategy with the sensing/computing. We call this general set of problems intermittent status updating. In this paper, we consider intermittent status updating through non-preemptive sensing/computing (S/C) and transmission (Tx) operations, each costing a single energy recharge of the node, through an erasure channel with perfect channel feedback and no channel feedback. The S/C time for each update is independent with a general distribution. The Tx queue has a single data buffer to save the latest packet generated after the S/C operation and a single transmitter where transmission time is deterministic. Once energy is harvested, the node has to decide whether to activate S/C to generate a new update or to (re)send the existing update (if any) to the receiver. We prove that when feedback is available average peak age of information (AoI) at the receiver is minimized by a threshold-based policy that allows only young packets to be (re)sent or else generates a new update. We additionally propose window based and probabilistic retransmission schemes and obtain closed form average peak AoI expressions. Our numerical results show average peak AoI performance comparisons and improvements.

Age of Information: An Indirect Way To Improve Control System Performance

Onur Ayan (Technical University of Munich, Germany); Anthony Ephremides (University of Maryland, USA); Wolfgang Kellerer (Technische Universität München, Germany)

In this paper, we consider N heterogeneous control sub-systems sharing a wireless communication channel. Network resources are limited and they are allocated by a centralized scheduler. Each transmission is lost with a probability that is higher or lower depending on the portion each sub-system receives from the pool of network resources. Furthermore, state measurements go through a first come first serve (FCFS) Geo/Geo/1 transmission queue after they are generated by each sensor. In such a setting, the information at each remote controller that is observing the state measurements through the wireless channel gets outdated. Age of Information (AoI) captures this effect and measures the information freshness at each controller. By definition, AoI is control unaware thus not a standalone metric to capture the heterogeneous requirements of control sub-systems. However, we show how the stationary distribution of Age of information (AoI) can be employed as an intermediate metric to obtain the expected control performance in the network. As a result, we solve the resource allocation problem optimally and show by simulations that we are able to improve the control performance indirectly through AoI.

Remote Tracking of Dynamic Sources under Sublinear Communication Costs

Jihyeon Yun (Korea University, Korea (South)); Atilla Eryilmaz (The Ohio State University, USA); Changhee Joo (Korea University, Korea (South))

We study the remote monitoring of multiple sensors with evolving states following a Wiener Process under communication cost. We assume that the communication cost is sublinear such that the cost decreases with the number of simultaneous state updates. Such sublinear structures emerge in various settings, such as frame aggregation, and give rise to interesting unexplored tradeoffs between: updating a smaller subset of the processes earlier at a higher cost-per-process; and updating a larger subset of them later at a lower cost-per-process. We attack this problem by first providing two competitive benchmark strategies of All-at-once and Multi-threshold policies. Then, we propose a novel strategy of MAX-k policy that not only includes the two benchmark threshold-based policies as special cases, but also improves over them by better exploiting the aforementioned tradeoff. Further, we develop the GPSO optimization technique to develop an online learning algorithm that adaptively optimizes the parameters of MAX-k policy. We demonstrate that the proposed scheme outperforms the well-known online learning algorithm based on UCB index.

Timely Tracking of Infection Status of Individuals in a Population

Melih Bastopcu (University of Maryland, USA); Sennur Ulukus (University of Maryland, USA)

We consider real-time timely tracking of infection status (e.g., covid-19) of individuals in a population. In this work, a health care provider wants to detect infected people as well as people who recovered from the disease as quickly as possible. In order to measure the timeliness of the tracking process, we use the long-term average difference between the actual infection status of the people and their real-time estimate by the health care provider based on the most recent test results. We first find an analytical expression for this average difference for given test rates, and given infection and recovery rates of people. Next, we propose an alternating minimization based algorithm to minimize this average difference. We observe that if the total test rate is limited, instead of testing all members of the population equally, only a portion of the population is tested based on their infection and recovery rates. We also observe that increasing the total test rate helps track the infection status better. In addition, an increased population size increases diversity of people with different infection and recovery rates, which may be exploited to spend testing capacity more efficiently, thereby improving the system performance. Finally, depending on the health care provider's preferences, test rate allocation can be altered to detect either the infected people or the recovered people more quickly.

Session AoI-S2

Session 2: AoI in Learning and Computing

11:00 AM — 12:50 PM EDT
May 10 Mon, 11:00 AM — 12:50 PM EDT

Timely Communication in Federated Learning

Baturalp Buyukates (University of Maryland, USA); Sennur Ulukus (University of Maryland, USA)

We consider a federated learning framework in which a parameter server (PS) trains a global model by using $n$ clients without actually storing the client data centrally at a cloud server. Focusing on a setting where the client datasets are fast changing and highly temporal in nature, we investigate the timeliness of model updates and propose a novel timely communication scheme. Under the proposed scheme, at each iteration, the PS waits for $m$ available clients and sends them the current model. Then, the PS uses the local updates of the earliest $k$ out of $m$ clients to update the global model at each iteration. We find the average age of information experienced by each client and numerically characterize the age-optimal $m$ and $k$ values for a given $n$. Our results indicate that, in addition to ensuring timeliness, the proposed communication scheme results in significantly smaller average iteration times compared to random client selection without hurting the convergence of the global learning task.

The Age of Correlated Features in Supervised Learning based Forecasting

Md Kamran Chowdhury Shisher (Auburn University, USA); Heyang Qin (University of Nevada, Reno, USA); Lei Yang (University of Nevada, Reno, USA); Feng Yan (University of Nevada, Reno, USA); Yin Sun (Auburn University, USA)

In this paper, we analyze the impact of information freshness on supervised learning based forecasting. In these applications, a neural network is trained to predict a time-varying target (e.g., solar power), based on multiple correlated features (e.g., temperature, humidity, and cloud coverage). The features are collected from different data sources and are subject to heterogeneous and time-varying ages. By using an information-theoretic approach, we prove that the minimum training loss is a function of the ages of the features, where the function is not always monotonic. However, if the empirical distribution of the training data is close to the distribution of a Markov chain, then the training loss is approximately a non-decreasing age function. Both the training loss and testing loss depict similar growth patterns as the age increases. An experiment on solar power prediction is conducted to validate our theory. Our theoretical and experimental results suggest that it is beneficial to (i) combine the training data with different age values into a large training dataset and jointly train the forecasting decisions for these age values, and (ii) feed the age value as a part of the input feature to the neural network.

Autonomous Maintenance in IoT Networks via AoI-driven Deep Reinforcement Learning

George J. Stamatakis (ICS-FORTH, Greece); Nikolaos Pappas (Linköping University, Sweden); Alexandros Fragkiadakis (Institute of Computer Science, FORTH, Greece); Apostolos Traganitis (University of Crete & ICS-FORTH, Greece)

Internet of Things (IoT) with its growing number of deployed devices and applications raises significant challenges for network maintenance procedures. In this work, we formulate a problem of autonomous maintenance in IoT networks as a Partially Observable Markov Decision Process. Subsequently, we utilize Deep Reinforcement Learning algorithms (DRL) to train agents that decide if a maintenance procedure is in order or not and, in the former case, the proper type of maintenance needed. To avoid wasting the scarce resources of IoT networks we utilize the Age-of-Information (AoI) metric as a reward signal for the training of the smart agents. AoI captures the freshness of the sensory data which are transmitted by the IoT sensors as part of their normal service provision. Numerical results indicate that AoI integrates enough information about the past and present states of the system to be successfully used in the training of smart agents for the autonomous maintenance of the network.

Computing-aided Update for Information Freshness in the Internet of Things

Minghao Fang (Sun Yat-Sen University, China); Xijun Wang (Sun Yat-sen University, China); Chao Xu (Northwest A&F University, China); Howard Yang (ZJU-UIUC Institute, China); Tony Q. S. Quek (Singapore University of Technology and Design, Singapore)

Age of information (AoI), a notion that measures the information freshness, is an important performance metric for real-time applications in Internet of Things (IoT). With the surge of computing resources at the IoT devices, it is possible to preprocess the information packets that contain the status update before sending them to the destination so as to lighten the transmission burden. However, the additional time and energy expenditure induced by computing also make the optimal updating a non-trivial problem. In this paper, we consider a real-time IoT monitoring system, where the computing-aided IoT device is capable of preprocessing the status update. A joint preprocessing and transmission policy is devised to minimize the average AoI at the destination and the energy consumption at the IoT device. Due to the difference in the processing rate and the transmission rate and the difference in the idle duration and the active duration, this problem is formulated as an average cost semi-Markov decision process (SMDP) and then transformed into a discrete-time Markov decision process (MDP). We show that the optimal policy is of threshold type with respect to the AoI. Equipped with this, a low-complexity relative policy iteration algorithm is proposed to obtain the optimal policy of the SMDP. Finally, simulation results demonstrate the optimal policy structure in different cases and show that the proposed policy outperforms two baseline policies.

Multi-UAV-enabled AoI-aware WPCN: A Multi-agent Reinforcement Learning Strategy

Omar Sami Oubbati (University of Laghouat, Algeria); Mohammed Atiquzzaman (University of Oklahoma, USA); Abderrahmane Lakas (UAE University, United Arab Emirates); Abdullah Baz (UQU, Saudi Arabia); Hosam Alhakami (Umm Al-Qura University, Saudi Arabia); Wajdi Alhakami (Taif University, Saudi Arabia)

Unmanned Aerial Vehicles (UAVs) have been deployed in virtually all tasks of enabling wireless powered communication networks (WPCNs). To ensure sustainable energy support and timely coverage of terrestrial Internet of Things (IoT) devices, a UAV needs to continuously hover and transmit wireless energy signals to charge these devices in the downlink. Then, the devices send their independent information to the UAV in the uplink. However, it was noted that the majority of existing schemes related to UAV-enabled WPCN are mainly based on a single UAV and cannot meet the requirements of a large-scale WPCN. In this paper, we design a separated UAV-assisted WPCN system, where two UAVs are deployed to behave as a UAV data collector (UAV-DC) and UAV energy transmitter (UAV-ET), respectively. Thus, the collection of fresh information and energy transfer are treated separately at the level of the two corresponding UAVs. These two tasks could be enhanced by optimizing the UAVs' trajectories. For this purpose, we leverage a multi-agent deep Q-network (MADQN) strategy to provide appropriate UAVs' trajectories that jointly minimize the expected age of information (AoI), enhance the energy transfer to devices, and minimize the energy consumption of UAVs. Simulation results show that our system enhances the performance of our strategy significantly in terms of AoI and energy transfer compared with baseline methods.

Session AoI-S3

Session 3: AoI in communications and networking

2:10 PM — 4:00 PM EDT
May 10 Mon, 2:10 PM — 4:00 PM EDT

Age Debt: A General Framework For Minimizing Age of Information

Vishrant Tripathi (MIT, USA); Eytan Modiano (MIT, USA)

We consider the problem of minimizing age of information in general single-hop and multihop wireless networks. First, we formulate a way to convert AoI optimization problems into equivalent network stability problems. Then, we propose a heuristic low complexity approach for achieving stability that can handle general network topologies; unicast, multicast and broadcast flows; interference constraints; link reliabilities; and AoI cost functions. We provide numerical results to show that our proposed algorithms behave as well as the best known scheduling and routing schemes available in the literature for a wide variety of network settings.

An Empirical Study of Ageing in the Cloud

Tanya Shreedhar (Indraprastha Institute of Information Technology (IIIT) Delhi, India); Sanjit K Kaul (IIIT Delhi, India); Roy Yates (Rutgers University, USA)

We quantify, over inter-continental paths, the ageing of TCP packets, throughput and delay for different TCP congestion control algorithms containing a mix of loss-based, delay-based and hybrid congestion control algorithms. In comparing these TCP variants to ACP+, an improvement over ACP, we shed better light on the ability of ACP+ to deliver timely updates over fat pipes and long paths. ACP+ estimates the network conditions on the end-to-end path and adapts the rate of status updates to minimize age. It achieves similar average age as the best (age wise) performing TCP algorithm but at end-to-end throughputs that are two orders of magnitude smaller. We also quantify the significant improvements that ACP+ brings to age control over a shared multiaccess channel.

Timely Transmissions Using Optimized Variable Length Coding

Ahmed Arafa (University of North Carolina at Charlotte, USA); Richard Wesel (University of California, Los Angeles, USA)

A status updating system is considered in which a variable length code is used to transmit messages to a receiver over a noisy channel. The goal is to optimize the codewords lengths such that successfully-decoded messages are timely. That is, such that the age-of-information (AoI) at the receiver is minimized. A hybrid ARQ (HARQ) scheme is employed, in which variable-length incremental redundancy (IR) bits are added to the originally-transmitted codeword until decoding is successful. With each decoding attempt, a non-zero processing delay is incurred. The optimal codewords lengths are analytically derived utilizing a sequential differential optimization (SDO) framework. The framework is general in that it only requires knowledge of an analytical expression of the positive feedback (ACK) probability as a function of the codeword length.

Can AoI and Delay be Minimized Simultaneously with Short-Packet Transmission?

Jie Cao (Harbin Institute of Technology, Shenzhen, China); Xu Zhu (University of Liverpool, United Kingdom (Great Britain) & Harbin Institute of Technology, Shenzhen, China); Yufei Jiang (Harbin Institute of Technology, Shenzhen, China); Zhongxiang Wei (University College London & School of EE and CS, United Kingdom (Great Britain))

Age of information (AoI) and delay are both critical performance metrics to enable emerging applications such as remote estimation and real-time control in factory automation. In particular, it remains an open question whether AoI and delay can be minimized simultaneously in short-packet transmission for factory automation etc., as they are affected by block length and update rate in a complex manner. Motivated by the open issue, we derive closed-form expressions and tight approximations for the average AoI and the average delay for a short-packet Last-Come First-Served system with the retransmission scheme and non-preemption policy. It is proved that there exists a strong tradeoff between AoI and delay with given block length, and that the two performance metrics can be minimized simultaneously with given status update rate. With the goal of minimizing delay and AoI simultaneously, the weighted sum of delay and AoI is formulated and proved to be convex with respect to block length and update rate, respectively. A low-complexity optimization algorithm is developed with the closed-form expression of the optimal update rate and optimal block length, whose performance approaches the Pareto boundary of the AoI-delay region.

Performance Analysis for Correlated AoI and Energy Efficiency in Heterogeneous CR-IoT System

Xiaoyu Hao (Fudan University, China); Tao Yang (Fudan University, China); Yulin Hu (RWTH Aachen University, Germany); Bo Hu (Fudan University, Shanghai, China)

We consider a cognitive radio based Internet of Things (CR-IoT) system where the secondary IoT device (SD) accesses the licensed channel during the transmission vacancies of the primary IoT device (PD). We focus on the impact of the IoT devices' heterogeneous traffic pattern on the energy efficiency and on the age of information (AoI) performance of the SD. We first derive closed-form expressions of the energy efficiency and the average AoI, and subsequently explore their convexity and monotonicity to the transmit power. Following these characterizations, an optimal transmit power optimization algorithm (TPOA) is proposed for the SD to maximize the energy efficiency while maintaining the average AoI under a predefined threshold. Numerical results verify the different preferences of the SD toward different PD traffic patterns, and provides insights into the tradeoff between the energy efficiency and the average AoI.

Made with in Toronto · Privacy Policy · INFOCOM 2020 · © 2021 Duetone Corp.