The full name of COS is Chip Operating System, which is generally developed around the characteristics of the smart cards it serves. Due to the inevitable impact of the performance and memory capacity of the microprocessor chip inside the smart card, COS is largely different from the operating systems we usually see on microcomputers (such as DOS, UNIX, etc.).

Firstly, COS is a specialized system rather than a universal system. That is to say, a type of COS can generally only be applied to a specific type (or certain types) of smart card, and the COS within different cards is generally different. Because COS is generally designed and developed specifically based on the characteristics and application scope of a certain smart card, although most of the functions they actually complete may follow the same international standard. Secondly, compared to common operating systems on microcomputers, COS is essentially closer to a temporary control program rather than a commonly referred to true operating system, which at least remains so at present. Because at the current stage, COS mainly needs to solve the problem of how to handle and respond to external commands, which generally does not involve the management and processing of sharing and concurrency. Moreover, considering the current application situation of smart cards, concurrency and sharing work are indeed not necessary. COS is generally designed and developed in close conjunction with the partitioning of smart card memory, following some of the functions specified in international standards (ISO/IEC 7816 series standards). However, due to the rapid development of smart cards and the relatively long development cycle of international standards, the current international standards for smart cards are not yet perfect. Therefore, many manufacturers have expanded their own developed COS.

At present, no company’s COS products can form an industrial standard. Therefore, this article will mainly combine existing (referring to pre 1994) international standards, focusing on the basic principles and functions of COS, and appropriately list their implementation methods in certain products as examples. The main function of COS is to control the information exchange between the smart card and the outside world, manage the memory inside the smart card, and complete various command processing within the card. Among them, exchanging information with the outside world is the most basic requirement of COS. During the exchange process, COS currently follows two types of information exchange protocols: the T=0 protocol for asynchronous character transmission and the T=l protocol for asynchronous packet transmission. The specific content and implementation mechanism of these two information exchange protocols are specified in the IS0/IEC78l6-3 and IS0/IEC7816-3A3 standards; The fundamental functions of management and control that COS should accomplish are specified in the ISO/IEC 78l6-4 standard. In this international standard, detailed explanations are also provided for the data structure of smart cards and the basic command set of COS. As for IS0/IEC78l6-l and 2, they specify the physical parameters and external dimensions of smart cards, and their relationship with COS is not very close.

In the early stages of IP network development, Internet applications (such as email, remote login, and web access) did not have special performance requirements, and a single Best Effort service class could support all Internet applications.

1. Receiving and sending emails is the earliest and most widely used network application. Due to its low cost and fast and convenient characteristics, it seems to shorten the spatial distance between people. Whether in a foreign country, exchanging information with friends or contacting work is as easy as chatting with neighbors next door. The concept of a global village is indeed reasonable.

The widespread application of the internet will create a digital way of life and work, called the SOHO (Small Home Office) approach. The family will no longer be just an isolated unit of human social life, but a vibrant cell in the information society.

3. Browsing or surfing the internet is the most basic service provided by the internet. You can visit any website online and freely explore the world according to your interests, without leaving your home.

4. Query information. Using the internet, the world’s largest database, you can use some search engines for searching information to find the information you need from a vast information database. With the development of the “government internet” project in our country, some daily things of people can be completely completed on the internet.

5. E-commerce is the behavior of consumers using the internet to enter online shopping sites for consumption. Online shopping sites are built in virtual digital spaces, utilizing the Web to showcase products and multimedia features to enhance their visibility and selectivity. Although online shopping is not yet perfect and will not replace traditional shopping methods, it is only a supplement to traditional shopping methods. But it has truly come to us, giving us an additional choice in our lives.

6. Enrich people’s leisure lifestyle. Leisure activities refer to non occupational labor activities, which include leisure and entertainment activities such as listening to music, watching movies, watching TV, dancing, and participating in sports activities; Developmental activities include learning cultural knowledge, participating in social activities, engaging in artistic creation and scientific invention activities, etc. But leisure life directly related to the internet generally includes leisure education, leisure entertainment, and leisure communication.

7. With the increasingly widespread application of internet pagers and other devices in people’s lives, everyone can make friends online from all over the world and exchange ideas with each other through the internet, truly achieving the goal of “having close friends in the sea, and the world is as close as a neighbor”.

8. Other applications. Online versions of human activities in the real world are ubiquitous, such as online streaming, stock trading, job hunting, art exhibitions, and more.

This problem is too difficult, it requires creativity, imagination, and the courage to practice. I only think of two words: engine. It may be an engine seeking development for various industries, people of all kinds, and all sectors of society, and it will bring human society into a brand new stage of digital history. The Internet has evolved into the driving force and new engine of the information economy, a bulldozer that reduces costs, improves productivity, and paves the way for various new jobs.

Because in the future, high-speed network connections and internet connections will be universal standards and ubiquitous. We will have incredible network connection speeds at home and in the office, and high-speed network connections will spread across various environments. The internet will completely constrain our work, life, and entertainment. I believe that the application of the internet will emerge like mushrooms after rain, constantly changing, just like the growth rate of internet users. With the global popularization of the Internet and its widespread application in various fields, the model of countries and enterprise groups in the industrial era, which relied on geopolitical site segmentation and monopoly, will gradually be broken. Facing a unified global market, the economy will achieve globalization. The most prominent economic model currently is e-commerce in the online environment.

But the huge commercial success of REC650 1MRK008514-CB has changed all of this. These changes are mainly reflected in the following aspects:

(1) Bandwidth is increasingly becoming a network bottleneck.

(2) The demand for network applicability and reliability is increasing.

(3) To stand out in competition, service providers must provide a variety of service types to adapt to different user groups.

(4) Service providers may even be required to provide services with stricter performance requirements, such as Voice over IP and Virtual Private Network (VPN), which traditional Internet cannot provide.

Based on the above changes, REC650 1MRK008514-CB service providers must provide multiple services on the same IP infrastructure, which have different performance requirements. In addition, with the popularization of data network applications, the demand for integrating traditional services (voice communication, video transmission, and data transmission) provided by three different networks into a single IP infrastructure is becoming increasingly urgent. This type of data network that provides multiple services on a single IP infrastructure is called a multi-service network, or a unified network. After integrating the network into a unified network, service providers can attract and retain more new users due to the wide range of services they provide. Moreover, the unified network effectively reduces the number of network devices that need to be managed, minimizes operating costs, and can provide users with high-quality and affordable services, thus attracting more users. Due to the use of packet switching technology in data networks, they have higher bandwidth utilization than traditional voice communication networks and video transmission networks.

There are currently two methods for implementing a unified network, and their implementation mechanisms are fundamentally different. One method is to exponentially increase network bandwidth by significantly increasing it to meet the performance requirements of a unified network; Another approach is to meet the performance requirements of a unified network by adding bandwidth management based service classification (CoS) functionality in the IP core network.

The main objections to the first method are as follows:

(1) The increase in cost and complexity of information flow management may offset the benefits of increasing bandwidth.

(2) In large IP networks, monitoring, authentication, and billing for various service types are very difficult.

Moreover, in the existing network infrastructure, there are already network infrastructures with Service Classification (CoS) capabilities, such as ATM networks, and these network infrastructures have been used to support services with strict performance requirements.

The main advantages of the second method are:

(1) Network management can optimize bandwidth utilization to the greatest extent possible, enabling the network to provide users with the maximum transmission capacity.

(2) When new applications emerge, service providers may need to adopt new network devices, which may increase the local bandwidth of the network. But these events rarely occur in a coordinated and consistent manner, which is highly likely to cause network bandwidth mismatch issues. Bandwidth management allows service providers to control bandwidth and handle the harm caused by bandwidth mismatch in paired networks.

(3) Bandwidth management adds value to the network by providing multiple services to users on the same network infrastructure. A unified network achieved through bandwidth management can reduce operating costs and provide users with multiple access services using a single access device, such as Internet access, VPN access, voice communication, and video conferencing.

The necessity of Service Classification (CoS)

A router is a packet switching device that achieves statistical multiplexing. And Internet Protocol (IP) is a datagram protocol that independently selects the transmission path (route) for each message, without the concept of connection. Traditional IP only provides a single type of best effort service. In this service mode, all packets transmitted through the network have the same priority, and doing their best means that the IP network will do everything possible to deliver the datagrams to their destination correctly and completely. However, it cannot guarantee that datagrams will not be discarded, damaged, duplicated, out of order, or sent incorrectly during transmission. In addition, it does not make any commitment to the transmission characteristics related to datagram transmission quality, such as transmission delay, delay jitter, throughput rate, etc.

If the network only satisfies customers through best effort service and does not make any judgments on discarded packets, network operations will inevitably encounter problems. The solution to this problem is to use intelligent transport layer protocols, and the only reason why best effort services can operate well in global IP networks is that the TCP protocol can monitor the occurrence of packet loss when the network is congested, and respond to packet loss by reducing transmission speed. It is precisely because of the role of TCP that the best effort service queue in routers becomes a queue with good behavior.

Best effort service is not a derogatory term. On the contrary, it is this ability to support a single best effort service that makes large IP networks and the Internet an irreplaceable technology today, and the only transmission platform that supports critical transaction applications on a global scale. Of course, the ability of IP networks to only support a single best effort service has also brought some problems and had a negative impact on its further commercial success. Many service providers have urgently felt the need to provide multiple service types in order to further support newly developed applications, so that each service type can meet specific performance requirements.

Providing different services for specific applications and users naturally involves the issue of differential treatment of messages belonging to different applications. Ultimately, multiple services are unfair services that provide unified management for specific information flows.

Service classification makes users feel that the quality of service (QoS) provided by the network for a specific data flow has changed by controlling some key attributes of the data flow. These key attributes include: (1) the amount of data transmitted per unit of time (throughput rate); (2) The time (delay) it takes to transfer data from one node in the network to another; (3) The range of transmission delay variation (delay jitter) between two nodes belonging to the same data flow; (4) The proportion of messages that cannot reach their destination correctly (loss rate).

It is worth noting that the quality of service for a specified service class depends on the quality of service provided by the worst performing link on the end-to-end transmission path.

It should be pointed out that reserving bandwidth for high priority services or reducing overload coefficients to support service classification will lead to a decrease in bandwidth utilization. In terms of network bandwidth utilization, a single best effort service achieved through statistical multiplexing has the highest utilization rate of network bandwidth. However, the loss suffered by service classification in terms of network bandwidth utilization is negligible compared to the value-added of network services caused by the support of newly developed applications such as voice transmission, video conferencing, and critical transaction applications.