Event

We study information freshness in cache updating systems, where users track a time-varying capacity-constrained source with the aid of network of cache-aided relay nodes. Two categories of networks are considered: multi-hop networks and gossip networks. We address open problems regarding characterization of the relevant timeliness metrics in these networks, deriving their closed form expressions for their long-term expected values, devising policies to achieve optimal freshness in these networks, and studying their robustness to various adversarial attacks.

In our first completed work, we consider timeliness of files in a parallel cache network, where we show that freshness for a file increases with increase in consolidation of rates across caches, and use it to find an approximate policy for assigning rates to all files. In our second completed work, we examine the resilience of an n node gossip network against jamming attacks as a function of its connectivity, by studying the average version age of the network in the presence of ñ jammers that disrupt inter-node communications, for the connectivity-constrained ring topology and the connectivity-rich fully connected topology. In our third completed work, we study the susceptibility of gossip networks to timestomping attacks, where an oblivious adversary infects the packets at a target node through data timestamp manipulation (or timestomping) to make fresh packets look stale and stale packets look fresh, with the intent to replace circulation of fresh packets with outdated packets in the network. We demonstrate how network topology capacitates an adversary to influence age scaling in a network, by studying both extremes of network connectivity spectrum, the fully connected network and bidirectional ring network.

Next, we present three proposed works. In our first proposed work, we consider a network consisting of n nodes that aim to track a continually updating process or event, such that to disseminate updates about the event to the network, m sources of varying degree of information reliability are available, and the network nodes prefer a more reliable packet over a less unreliable packet even when the former is a bit outdated with respect to the latter. We aim to study how such preference affects the fraction of users with reliable information in the network and their version age of information. In our second proposed work, we consider cache updating systems, where updates on all links are forwarded according to ordinary renewal processes that are not necessarily Poisson processes. The evolution of age in the given network is defined by a superposition of multiple independent ordinary renewal processes and it becomes difficult to calibrate the different renewal processes with respect to each other, making analysis challenging. In this regard, we aim to characterize the different timeliness metrics in arbitrary networks. In our third proposed work, we aim to improve network freshness by speeding up dissemination of k time varying files in a gossip network of n nodes through random linear coding.