Latency (engineering)

About: Latency (engineering) is a research topic. Over the lifetime, 7278 publications have been published within this topic receiving 115409 citations. The topic is also known as: lag.

Fault tolerant scheduling of precedence task graphs on heterogeneous platforms

Abstract: Fault tolerance and latency are important requirements in several applications which are time critical in nature: such applications require guaranties in terms of latency, even when processors are subject to failures. In this paper, we propose a fault tolerant scheduling heuristic for mapping precedence task graphs on heterogeneous systems. Our approach is based on an active replication scheme, capable of supporting epsiv arbitrary fail-silent (fail-stop) processor failures, hence valid results will be provided even if epsiv processors fail. We focus on a bi-criteria approach, where we aim at minimizing the latency given a fixed number of failures supported in the system, or the other way round. Major achievements include a low complexity, and a drastic reduction of the number of additional communications induced by the replication mechanism. Experimental results demonstrate that our heuristics, despite their lower complexity, outperform their direct competitor, the FTBAR scheduling algorithm [3].

Broadband Analog Aggregation for Low-Latency Federated Edge Learning

Abstract: The popularity of mobile devices results in the availability of enormous data and computational resources at the network edge. To leverage the data and resources, a new machine learning paradigm, called edge learning, has emerged where learning algorithms are deployed at the edge for providing fast and intelligent services to mobile users. While computing speeds are advancing rapidly, the communication latency is becoming the bottleneck of fast edge learning. To address this issue, this work is focused on designing a low latency multi-access scheme for edge learning. We consider a popular framework, federated edge learning (FEEL), where edge-server and on-device learning are synchronized to train a model without violating user-data privacy. It is proposed that model updates simultaneously transmitted by devices over broadband channels should be analog aggregated "over-the-air" by exploiting the superposition property of a multi-access channel. Thereby, "interference" is harnessed to provide fast implementation of the model aggregation. This results in dramatical latency reduction compared with the traditional orthogonal access (i.e., OFDMA). In this work, the performance of FEEL is characterized targeting a single-cell random network. First, due to power alignment between devices as required for aggregation, a fundamental tradeoff is shown to exist between the update-reliability and the expected update-truncation ratio. This motivates the design of an opportunistic scheduling scheme for FEEL that selects devices within a distance threshold. This scheme is shown using real datasets to yield satisfactory learning performance in the presence of high mobility. Second, both the multi-access latency of the proposed analog aggregation and the OFDMA scheme are analyzed. Their ratio, which quantifies the latency reduction of the former, is proved to scale almost linearly with device population.

Overhead and performance of low latency live streaming using MPEG-DASH

Abstract: HTTP Streaming is a recent topic in multimedia communications with on-going standardization activities, especially with the MPEG DASH standard which covers on demand and live services. One of the main issues in live services deployment is the reduction of the overall latency. Low or very low latency streaming is still a challenge. In this paper, we push the use of DASH to its limits with regards to latency, down to fragments being only one frame, and evaluate the overhead introduced by that approach and the combination of: low latency video coding techniques, in particular Gradual Decoding Refresh; low latency HTTP streaming, in particular using chunked-transfer encoding; and associated ISOBMF packaging. We experiment DASH streaming using these techniques in local networks to measure the actual end-to-end latency, as low as 240 milliseconds, for an encoding and packaging overhead in the order of 13% for HD sequences and thus validate the feasibility of very low latency DASH live streaming in local networks.

Inhibition of NF-κB Activation In Vivo Impairs Establishment of Gammaherpesvirus Latency

Abstract: A critical determinant in chronic gammaherpesvirus infections is the ability of these viruses to establish latency in a lymphocyte reservoir. The nuclear factor (NF)-κB family of transcription factors represent key players in B-cell biology and are targeted by gammaherpesviruses to promote host cell survival, proliferation, and transformation. However, the role of NF-κB signaling in the establishment of latency in vivo has not been addressed. Here we report the generation and in vivo characterization of a recombinant murine gammaherpesvirus 68 (γHV68) that expresses a constitutively active form of the NF-κB inhibitor, IκBαM. Inhibition of NF-κB signaling upon infection with γHV68-IκBαM did not affect lytic replication in cell culture or in the lung following intranasal inoculation. However, there was a substantial decrease in the frequency of latently infected lymphocytes in the lung (90% reduction) and spleens (97% reduction) 16 d post intranasal inoculation. Importantly, the defect in establishment of latency in lung B cells could not be overcome by increasing the dose of virus 100-fold. The observed decrease in establishment of viral latency correlated with a loss of activated, CD69hi B cells in both the lungs and spleen at day 16 postinfection, which was not apparent by 6 wk postinfection. Constitutive expression of Bcl-2 in B cells did not rescue the defect in the establishment of latency observed with γHV68-IκBαM, indicating that NF-κB–mediated functions apart from Bcl-2–mediated B-cell survival are critical for the efficient establishment of gammaherpesvirus latency in vivo. In contrast to the results obtained following intranasal inoculation, infection of mice with γHV68-IκBαM by the intraperitoneal route had only a modest impact on splenic latency, suggesting that route of inoculation may alter requirements for establishment of virus latency in B cells. Finally, analyses of the pathogenesis of γHV68-IκBαM provides evidence that NF-κB signaling plays an important role during multiple stages of γHV68 infection in vivo and, as such, represents a key host regulatory pathway that is likely manipulated by the virus to establish latency in B cells.