In multi-access edge computing (MEC), most existing task software caching works focus on statically data at the network edge, which may hardly preserve high reusability due to time-varying user requests in practice. To this end, work considers dynamic MEC server assist users' execution. Specifically, we formulate a joint update (TSCU) and computation offloading (COMO) problem minimize energy consumption while guaranteeing delay constraints, where limited cache size capability of server, as...

At the network edges, multi-tier computing framework provides mobile users with efficient cloud-like and signal processing capabilities. Deploying digital twins in system helps to realize ultra-reliable low-latency interactions between their virtual objects. Considering may roam edge servers limited coverage increase data synchronization latency twins, it is crucial address twin migration problem enable real-time users. To this end, we formulate a joint migration, communication computation...

Utilizing the data caching technology to reduce transmission is a promising technique for improving performance of mobile edge computing (MEC), because delay and energy consumption produced by constitute dominant cost task execution in MEC. Besides, computation tasks generally consist input parameters, executive codes, results. The codes are fixed can output difference results under different parameters. Motivated this, we consider proactively cache at MEC server weighted sum users'...

The conventional model aggregation-based federated learning (FL) approach requires all local models to have the same architecture, which fails support practical scenarios with heterogeneous models. Moreover, frequent exchange is costly for resource-limited wireless networks since modern deep neural usually over a million parameters. To tackle these challenges, we first propose novel knowledge-aided FL (KFL) framework, aggregates light high-level data features, namely knowledge, in per-round...

Federated learning (FL) is an efficient and privacy-preserving distributed paradigm that enables massive edge devices to train machine models collaboratively. Although various communication schemes have been proposed expedite the FL process in resource-limited wireless networks, unreliable nature of channels was less explored. In this work, we propose a novel framework, namely with gradient recycling (FL-GR), which recycles historical gradients unscheduled transmission-failure improve...

Existing device scheduling works in wireless federated learning (FL) mainly focused on selecting the devices with maximum gradient norm or loss function and require all to perform local training each round. This may produce extra costs schedule similar data statistics, thus degrading performance. To mitigate these problems, we first theoretically characterize convergence behaviour of considered FL system, finding that performance is degraded by difference between aggregated scheduled full...

Recently, the incidence of adenocarcinoma esophagogastric junction (AEG) is increasing in China. Laparoscopic total gastrectomy plus D2 lymph node dissection an important treatment method for AEG Siewert type II. In order to achieve esophagojejunostomy, surgical techniques, no matter open surgery or laparoscopic one, are highly demanding. This report describes a case complication during gastrectomy, which pseudoaneurysm developed at gastroesophageal anastomosis. The condition was...

The data heterogeneity across clients and the limited communication resources, e.g., bandwidth energy, are two of main bottlenecks for wireless federated learning (FL). To tackle these challenges, we first devise a novel FL framework with partial model aggregation (PMA). This approach aggregates lower layers neural networks, responsible feature extraction, at parameter server while keeping upper layers, complex pattern recognition, personalization. proposed PMA-FL is able to address reduce...

Most existing wireless federated learning (FL) studies focused on homogeneous model settings where devices train identical local models. In this setting, the with poor communication and computation capabilities may delay global update degrade performance of FL. Moreover, in homogenous settings, scale is restricted by device lowest capability. To tackle these challenges, work proposes an adaptive pruning-based FL (AMP-FL) framework, edge server dynamically generates sub-models pruning for...

Purpose: To evaluate the predictive capacity of nutritional-inflammatory index and clinicopathological characteristics in patients with locally advanced rectal cancer (LARC) receiving total neoadjuvant therapy (TNT). Methods: Data from 127 LARC TNT January 2017 to 2021 were retrospectively analyzed. Clinicopathological different TNT-induced responses compared. The Chi-square test Mann–Whitney used analyze association between pre-TNT factors responses. Multivariable logistic regression...

Applications with more sensitive delay and larger data volumes, such as interactive gaming augmented reality, have become popular recently. Computation offloading is expected a promising technique to meet low latency for mobile users. However, computation requires communication between users the edge computing (MEC) server, energy consumption caused by transmission are considerable expenses Motivated this, we consider joint task caching optimization in cellular network where can proactively...

Caching popular contents on mobile devices is a promising technique to alleviate the backhaul data rate requirements. Since both file placement and exchange among consume energy, energy status of device can have significant impact caching utility whole system. This work considers optimization in cellular network, where are charged with harvested from ambient environment. As collect segments local storage via device-to-device (D2D) links links, we aim at minimizing percentage segment that...

Mobile edge computing (MEC) provides information technology and cloud-computing capabilities within the network close to mobile users, thereby addressing demand for users. However, energy consumption of data uploading from users MEC server makes it hard meet users' in some specific applications, i.e., interactive gaming augmented reality. Motivated by this, we integrate a task caching mechanism into computation offloading technique. Specifically, allows proactively cache tasks offload their...

Applications with sensitive delay and sizeable data volumes, such as interactive gaming augmented reality, have become popular in recent years. These applications pose a huge challenge for mobile users limited resources. Computation offloading is mainstream technique to reduce execution save energy users. However, computation requires communication between edge computing (MEC) servers. Such mechanism would difficultly meet users' demand some data-hungry computation-intensive because the...

Preventive ileostomy (PI) is conventionally performed to prevent anastomotic leakage (AL) after laparoscopic total mesorectal excision (LTME) for low rectal cancer; however, secondary surgery required remove the ostomy tube. We designed a new type of ostomy, transcecum catheterization (TCI) AL. Its principle similar PI, but no operation needed. evaluated safety and efficacy TCI in AL prevention.We analyzed data patients who underwent LTME with anastomosis Chongqing University Cancer Hospital...

Cooperative caching has emerged as an efficient way to alleviate backhaul traffic and enhance user experience by proactively prefetching popular videos at the network edge. However, it is challenging achieve optimal design of video caching, sharing, delivery within storage-limited edge networks due growing diversity videos, unpredictable requirements, dynamic preferences. To address this challenge, work explores cost-efficient cooperative via compression techniques while considering unknown...

As an important construction material, the mechanical and deformation properties of marine coral sand determine safety stability related island coastal engineering construction. The porous easily broken characteristics often make it difficult to meet needs. In particular, undergoes a large amount particle breakage under high-stress conditions, which in turn negatively affects its properties. this study, macro- micro-mechanical behavior geosynthetic-reinforced high confining pressure was...