Masquerading a Networked Embedded Computer as a Remote Object Using an Asynchronous Remote Object Protocol for a System With Small Servers

Noriichi Kanaya

doi:10.1109/TIM.2023.3265764

Return

Masquerading a Networked Embedded Computer as a Remote Object Using an Asynchronous Remote Object Protocol for a System With Small Servers

2023-01-01

PRE

DOI：10.1109/TIM.2023.3265764

Original

Original request for help

Save

Abstract

En 中文

There are increasing demands for making multiple sensors/devices accessible over a wireless local area network (WLAN) for measurement, control, and monitoring purposes. Such a system comprises a client and multiple small embedded computer-based servers deployed in remote places. To apply the advantage of distributed object models, such as Java Remote Method Invocation (RMI), the model cannot work on an embedded computer because it demands sufficient resources, including the processing power of the central processing unit, memory, and an operating system. The model provides synchronous operations using the Transmission Control Protocol/Internet Protocol with heavy overheads, which results in low throughput compared with the lightweight User Datagram Protocol. A simple asynchronous remote object protocol for embedded computers (eASROP) is designed and implemented to enable an embedded computer and its acquired data to masquerade as an RMI-compatible remote object on a server. The client is not blocked after it makes a remote method call to a server and can continue its job. The client can asynchronously receive a return object that contains status data and control data. The results of performance measurements obtained in experiments demonstrated that, for the eASROP on an embedded computer-based server (48 MHz), the extra execution time that the client experienced was 1.67%-15.4% of that for the RMI on a personal-computer-based server (3.1 GHz, four cores) on a WLAN.

Keywords:

Servers

TCPIP

Java

Protocols

Distributed databases

Delays

Wireless LAN

Asynchronous distributed object model

control and monitoring system

data acquisition system

distributed measurement