Abstract
People have higher expectations for embedded systems as the Internet becomes more pervasive and communication technology develops quickly. Instead of only personal computers, they wish to use Internet applications on embedded systems connected to IP networks. Additionally, they desire Internet connectivity for various kinds of embedded devices. This paper compares and contrasts two embedded systems research articles. They anticipate that embedded systems will be able to employ IP-based Internet applications and that numerous embedded devices, in addition to PCs, will be able to connect to the Internet. The initial investigation by Lin et al. (2011) focuses on improving security and reliability in multi-core virtualization designs through hardware help, all while maintaining performance and eliminating the need for additional hardware or complex security mechanisms in the virtualization layer. In their second publication, Guan et al.(2008) introduce a monitoring system for embedded devices based on embedded internet technologies. The key theories, design strategies, implementations, discussions, connected works, and conclusions of both articles are summarized in this report.
Introduction
Embedded systems are receiving a lot of attention due to their expanding use in various sectors, including industrial automation, medical technology, automotive systems, and consumer electronics. Designers and consumers of embedded systems desire rich graphic support, dependable, off-the-shelf device drivers with many capabilities, etc. The virtualization technique addresses the reliability, complexity, and adequate handling of real-time task problems. Because of their distinctive qualities, embedded systems are well-suited for various applications. These characteristics include low power consumption, a tiny form factor, real-time operation, and reliability. However, designing and implementing reliable embedded systems present significant challenges due to several factors, including hardware constraints, software complexity, and real-time constraints. This essay reviews two studies that touch on some of these issues. In particular, the requirement for systems with real-time capabilities and dependability is covered in the first section, which also examines the growing need for embedded products. But due to these complex requirements, a single RTOS is no longer adequate, and GPOS like Linux cannot adequately handle real-time tasks.
On the other hand, a new virtualization architecture for embedded systems with multi-core processors has been developed. Its features include having the virtualization layer and the kernel of guest systems in kernel space, using a distributed virtualization layer design, and ensuring isolation through local memory. In multi-core embedded systems, local memory and virtualization architecture can boost security and reliability, and EI-based remote monitoring systems can boost productivity and cut operating costs. The paper will go into greater detail on the advantages and difficulties of virtualized embedded systems.
Main Theory
An improvement in the reliability of multi-core virtualized embedded systems is suggested in the first study by Lin et al. (2011). The authors address the difficulties brought on by the growing need for dependable embedded goods and extensive graphic support. They recommend utilizing a hardware-based fault detection and recovery technique that divides physical hardware into many instances, each controlled and monitored by a separate Virtual Machine Monitor (VMM). The technique combines a fault detector and a fault handler to detect and recover from various forms of hardware and software defects. The monitoring system used by Guan et al.(2008) to monitor and manage embedded devices is presented in the second publication. An embedded monitoring module and a monitoring server are the two components of the suggested system. The authors explain how the system uses Embedded Web Server (EWS) technology to embed the EWS in the device and adopts the Browser/Server mode. This method allows embedded devices to be accessed, managed, and controlled from anywhere in the world using a regular Web browser. The authors show its success by using the suggested approach to temperature and humidity monitoring equipment.
Design Approaches
The hardware-based fault detection and recovery technique utilized in the first publication by Lin et al.(2011) is the design strategy. The hardware design includes a fault detector and a fault handler to detect and recover from different kinds of hardware and software problems. The paper describes a virtualization layer for embedded systems dubbed SPUMONE (Software Processing Unit, Multiplexing ONE into two or more). To lessen the effect of virtualization on guest operating systems, SPUMONE is a lightweight virtualization environment that can be run on a single or multi-core platform. The main attributes of SPUMONE are highlighted in the article, including its low emulation overhead, little need for guest OS change, interrupt virtualization, and distributed design for improved isolation and dependability. The distributed design increases the system’s stability, which assigns each SPUMONE instance to a physical core’s local memory space. The study compares the lines of code (LOC) changes made to guest operating systems and examines the implementation of SPUMONE on the Renesas RP1 processor, which has four physical cores.
Furthermore, An embedded monitoring system based on internet technology is the design strategy employed in the second study by Guan et al.(2008). Two components comprise the system: a monitoring server and an embedded monitoring module. The monitoring module gathers and transmits data online to the monitoring server. Web browser and EWS are the two components that make up the system, which uses a Browser/Server architecture. An Embedded Web Server (EWS) must be included in the device for internet connectivity. By adopting a hierarchical protocol stack made up of interactive modules, the system uses a simplified TCP/IP protocol stack, which minimizes the complexity of the conventional TCP/IP protocol stack, for storing monitoring web pages that the EWS accesses and other binary files that the system requires, the EWS offers an embedded file system, a compact file system based on an external serial flash chip.
Implementations
Lin et al. (2011) used the proposed hardware-assisted reliability boost on a multi-core virtualization system and then tested it with the MSRP1BASE2 board. Each core of the board’s RP1 quad-core 400MHz processor has its own ILRAM, OLRAM, and URAM of 8 KBytes, 16 KBytes, and 128 KBytes, respectively. The board also includes 128MB of memory. The authors ran SPUMONE instances with Linux 2.6.16 as the guest GPOS and TOPPERS/JSP 1.3 as the guest RTOS for the evaluation. The authors compared the differences in interrupt dispatch delay between the suggested technique and the initial RTOS implementation of SPUMONE. The measurement outcomes demonstrated that the dispatch latency distribution of the local memory implementation was almost identical to that of the original SPUMONE implementation. The original SPUMONE has an average dispatch delay of 8.54 us, compared to 7.1 us for the local memory implementation. The backbench benchmark application measured the Linux scheduler’s latency and performance differences when RTOS was executing its regular task. Due to platform limitations, this evaluation employed up to 10 groups.
Guan et al. (2008) created an Early Warning System (EWS) built within the gadget and gave it internet connectivity to monitor temperature and humidity sensors. The hardware interface, device gateway, embedded file system, and a thin TCP/IP protocol stack make up the EWS software architecture. When a web browser requests a URL, the EWS displays relevant monitoring web pages kept in the embedded file system. The embedded device gateway translates TCP/IP protocol into a customized device monitoring protocol for two-way communication and processes. It sends monitoring commands to the target device that are received from users via websites. Interactive elements like the application, transport, network, data connection, and physical layer make up the thin TCP/IP protocol stack. A file declaration that contains the filename, length, and start address makes it simple to include files in applications thanks to the system’s built-in file system. With these characteristics, the system offers a dependable way to operate and keep an eye on temperature and humidity monitoring equipment.
Discussion
The first publication by Lin et al.(2011) analyzes the suggested approach, noting its distinctive qualities while contrasting it with other approaches already in use. The capacity of local memory in embedded systems, which might need more to store a complete guest system, is one issue they raise. They recommend deciding whether vital components of the guest system should be sufficiently protected and retained in local memory. The authors also suggest a technique for adding a core-local memory protection mechanism to multi-core embedded systems to increase their security and dependability. This requires using a monitoring application to conduct data hashing and allocating certain portions of the guest system to local RAM as needed. Similarly, the second paper by Guan et al.(2008) describes the proposed EWS system, its benefits, and drawbacks, contrasts it with other current systems, and emphasizes its distinguishing characteristics. The remote monitoring system for Web UPS devices created for KYLINK Communications CORP is one use of the EWS. System users can use an internet web browser to access and control the UPS’s state data. The system’s performance was examined by running two processes on a PC. The testing outcomes show that the system can process UDP datagrams and react to HTTP requests in milliseconds, more than meeting the requirements of most EI applications.
Related Work
Better performance and less power consumption may result from the convention of local memory in embedded systems. The system’s performance can be improved by allocating programs to on-chip or local memory. Power consumption can be reduced by deciding which data should be stored in local memory. Furthermore, local memory offers increased isolation and security because only the processor core can access it. This can improve the dependability and security of embedded virtualization systems.
Conclusion & future work
Based on the two texts, using local memory and a new virtualization architecture for embedded systems powered by multi-core processors can improve system dependability and data security without adding overhead. However, a multi-core embedded platform still needs to handle the power consumption problem. Future work should concentrate on thoroughly analyzing the design. The second text by Guan et al.(2008) is designing and implementing a remote monitoring system built on EI technology for embedded devices. The technology links embedded devices to the Internet in Browser/Server mode, enabling remote users to view and control the devices using a regular web browser. The system can reduce operating costs and increase maintenance effectiveness, and it has been demonstrated to meet the requirements of EI applications.
References
Lin, T. H., Kinebuchi, Y., Courbot, A., Shimada, H., Morita, T., Mitake, H., … & Nakajima, T. (2011, March). Hardware-assisted reliability enhancement for embedded multi-core virtualization design. In 2011 14th IEEE International Symposium on Object/Component/Service-Oriented Real-Time Distributed Computing (pp. 241-249). IEEE.
Guan, M., Wei, W., & Bao, Y. (2008, December). A Monitoring System Based on Embedded Internet Technology for Embedded Devices. In 2008 International Conference on Computer Science and Software Engineering (Vol. 4, pp. 5-8). IEEE.
write