Artificial Intelligence and MCS Innovation

This document intends to analyze AI impact on MCS Innovation and assess the possibility of opening up a research field in AG 4 to study AI enabled MCS Innovation. As advisory group for SC 6, AG 4 should first understand what has been done in SC 6 in AI related research to avoid duplication/contradictions and to make better use of existing resources.

The research of AI in SC 6 started with the concept of AIEN “Artificial Intelligence Enabled Networking” in 2019. It is a research project in SC 6 WG 7. It is also the only such project so far in SC 6. It has a long history of development and had gone through several major changes in concept.

The following table provides a highlight of the timeline of AIEN history:

Above timeline can be categorized into four major stages

The first stage is from 2007 to 2014. During this seven year period, SC 6 founded Future Network project and completed the first phase of FN standardization by publish seven Technical Reports focusing on problem statement and requirements.

The second stage is from 2014 to 2017 in which ISO/IEC TR 29181-8 (FNQoS) was added and completed.

The third stage is from 2018 to 2023 in which FNQoS architecture and protocols (ISO/IEC 21558-2 and 21559-2) were developed and published.

The fourth stage is from 2019 to current which is almost parallel with the third stage and the scope has been extended.

Among the five new AIEN proposals, WG 7 N 387 is related to AG 4 and needs special attention. The title of WG 7 N 387 is “Harmonious collaboration between base station modulation and user applications”. here is a summary of the proposal:

In wireless communications, spectral utility and economic considerations require new technologies.

In short, WG 7 N 387 is offering a new standard which is AIEN based modulation technology with artificial intelligent enabled functions to reduce over service and other benefits.

In many ways, WG 7 N 387 is closed related to AG 4 and requires attention.

WG 7 N 387 is the first document connecting AIEN with modulation. Its usage in the SC6 scope is universal.. Therefore, it needs help and support from AG 4 which is a dedicated group for Modulation and Coding Scheme innovation standardization research.

AG 4 has another useful resource. WG 7 N 387 is proposing modulation technology based on AIEN which is also based on published standard on Future Network. AG 4 convener Kingston ZHANG is one of the veteran expert of ISO/IEC Future Network standardization. He was an active participant and made numerous contributions to SC 6 Future Network work and is one of the editors of ISO/IEC TR 29181-2 (Future Network Naming and Addressing).

A third advantage for AG 4 is its position. In SC 6, WG 1 is responsible for MAC and PHY layer standardization (communication) and WG 7 is specializing in Network layer standardization. Usually, modulation is considered a PHY layer technology. AG 4 is an SC 6 advisory group and therefore it is not restricted by the layer restrictions. AG 4 identifies and develops MCS technologies and can recommend standardization proposals to any SC 6 working groups.

WG 7 N 387 is interesting to AG 4 also because it opens up a direction to work on “modulation for Networking, MFN”.

Modulation is a PHY technology and its primary function is to facilitate communication. But when the communication equipment such as routers and switches is interconnected to form various networks, modulation also becomes the base technology for networking. The capabilities of modulation schemes will have direct impact on the performances of networks. Therefore, after a new MCS technology is identified and adopted as a SC 6 standard, apply it into networking standards can be done in parallel with works in communication.

In short, modulation technology is a work in PHY layer and primary application is to enhance communication efficiencies, but the work can be extended to cover networking. On the other hand, networking standard development may also find new requirement for enhanced Modulation services which is the case of WG 7 N 387.

Since N387 is relevant to AG 4's work, some procedural rules need to be established for coordination and mutual support. AG 4 needs to establish positions on a few issues which will not only affect WG 7 N 387 but also AG 4's work.

The issues include:

Should Artificial Intelligence be included in MCS Innovation research?

AIEN is the first basic standard in SC 6 to apply AI technology to modulation. It can guide the usage of AIEN. SC 6 needs to take a broader survey of the academics and technologies to make the proposals more inclusive.

Before the emergence of AI as a popular technology, ICT sector has long history of applying intelligent operation into communication and networking operations. Typical applications include channel and frequency selections, automatic adjustment of transmitting power levels, and automatic hibernation in no use situations.

However, traditional concept of automation is not complicated. It is intelligent but not at the advanced level. The trend is moving from being intelligent to highly advanced Artificial Intelligence which can utilize deep learning and logical reasoning to perform more complicated duties. The conceptual evolution from FNProxy in Future Network to AIEN in PWI-5096 reflect this trend.

It is clear that AI can be applied to make MCS operations more efficiently. As to whether AI can help AG 4 find innovative MCS solutions to overcome the theoretic and physical berreier, it needs more assessment.

Another direction is the emergence of Semantic Communication concept in the past decade which is hoping to rely on AI/DL technology to allow communication systems avoid the constraints of classic information theories. This direction has not been analyzed in AG 4 but should be given attention because AG 4 does have the resources to evaluate the technical feasibility and future impact of this technology.

The semantics of the three elements (syntax, semantics and timing) of communication protocol can be represented by proper semantic coding, which can improve the communication effect and avoid the restriction of Shannon's law.

But one problem with semantic communication approach is that sending less information could lead to misunderstanding or misinterpretation on the receiving end. Proponent of this direction points to the growth of AI/DL (Artificial intelligence/Deep Learning) as the future tool to deal with this problem.

Discussion on Semantic Communication and AIEN leads to a concept of Artificial Intelligence Enabled Communication (AIEC). The proxy mechanism in Future Network leads to the awareness that more intelligent solutions are needed for networking, the semantic communication technology also indicates the need for more intelligent solutions to improve communication services. This technical direction will continue to attract interests from the industry. The only uncertainty is about to what degree can AI impact communication performances.

There will be a race between AI and human wisdom to decide who can break the theoretical and physical limits in communication channel capacity.

RIS is the abbreviated term for Re-configurable intelligent surfaces, a new material based technology that is seen as have many benefits for enhanced communication and networking performances. The Surface material is capable of receiving, re-configuring, enhancing, and redirecting received radio frequency signals. Due to its potential to enhance the capacity and coverage of wireless networks by smartly re-configuring the wireless propagation environment, RIS is considered a promising technology for the sixth-generation (6G) of mobile communication, NFC and other wireless communication systems.

On 14 Dec. 2023, SC 6 circulated a Chinese contribution providing the first explanation on how RIS can enhance wireless communication and networking. The document introduced a new generation of RIS technology with abbreviated term of “MF-RIS”. Each element of the MF-RIS is composed of an amplifier and a phase-shifter, which can be used to achieve signal amplification, reflection and refraction. The MF-RIS first amplifies the incident signals, and then divides them into reflection and refraction parts by power split circuits. Thus, all users around the MF-RIS can be served simultaneously.

Clearly, RIS technology is intelligent and will most likely needs the support of Artificial Intelligence. On the other hand, from AG 4's perspective, RIS is capable of providing better and enhanced spectral utility rates and enhance user end efficiencies. Whether and how the technology can enhance MCS efficiencies or even provide an alternative direction different from AG 4's directions is unclear and is worthy more attention.

How AI can be used to promote MCS Innovation has not been studied in AG 4, but some positions can be stated:

Traditional automation control technologies can help communication and networking become intelligent and is useful in modulation and coding scheme services. But usually the technology is used to make micro-management of available resources.

The discussion on AIEN is beneficial because it raised an overlooked perspective on MCS Innovation. In AG 4 research so far, MCS Innovation are exclusively treated as a communication technology. WG 7 N 387 revealed a situation that networking also needs the support of advanced modulation technologies. MCS Innovation thus needs to consider the gaps and requirements of MCS operations in the perspective of networking.

Before discussion the relationship between MCS innovation and Networking, the relationship between MCS Innovation and communication needs to be first explained.

In SC 6, there is no definition for “communication.” A simple definition of communication is that it is a process of exchanging information between two entities. In ICT sector, communication is also known as telecommunication as shown in JTC 1/SC 6 title. Communication can take many forms and mediums such as gesture, voice, music, writing, drawing, etc. Adding prefix “tele” to communication indicates that the exchange of information can take place in long distances.

Communication requires at least two entities, one to send and the other to receive information. MCS is a key technology involving both end offering modulation and demodulation, coding and decoding of messages.

Because the MCS modulation-demodulation and coding-decoding procedures take place below the MAC layer it is designed as PHY layer technology and belongs to SC 6/WG scope.

According to SC 6 Business Plan, the work of WG1 is closely related to all sorts of physical communication technologies.

SC 6/WG 1 is to deliver the standards on services and protocols in the physical and data link layers. As new types of communication technologies emerge, it extends the working boundary and delivers the standards the market requires.

MCS Innovation technical objectives have not been clearly defined, but based on previous SC 6 and AG 4 documents, the following principles should be established:

Rule 1: MCS Innovation's objective is to find a way for long term advancement and improvement for communication efficiency.

The definition of “network” can be found in ISO/IEC standards including:

A collection of interconnected elements that provides connection services to users (ISO 1745:1975)

If compared with “communication”, networking can be viewed as an advanced form of communication. Networks can not only provide service to one to one communication, but also one to many communication (broadcasting) and many to many communication (meetings). Because of the more diversified application scenarios, networking is more complicated and sophisticated information system.

In ISO/IEC, network standardization is the duty of JTC 1/SC 6 whose title includes “information exchange between systems” which is another way to refer to “networking”. In SC 6, the work of networking standardization is the responsibility of WG 7 whose title is described as “Network, Transport, and Future Network”.

WG 7 had developed lots of protocols on OSI network and transport by the early 1990s. Currently, WG 7 has been working on Future Network including, other emerging networking issues, to prepare the future of the networks. Considering recent market requirements and activities of other SDOs on 5G/6G network technologies, WG 7 plans to challenge this important technical area as well as future networking technologies including multi-access edge computing, communication protocols in low earth orbit inter-satellite networks and application of AI to networks to support emerging new applications and services such as Blockchain, and metaverse.

The challenge is that almost all new emerging application and services have dedicated committees to handle International Standardization. WG 7 needs to possess unique resource and advantage to attract interest in seeking support for networking standardization. In this perspective, MCS Innovation could possibly be the new driving force.

On this subject, there are a few issues needing clarification and understanding:

Modulation and Coding Schemes are considered PHY layer protocols.

WG 7 N 387 has several definitions related to Modulation and Coding Scheme efficiencies. Some of the terms are also used in AG 4 technical research documents, but the meanings are different. AG 4 has not started offering terms and definitions in MCS Innovation standardization. The analysis below indicates that AG 4 needs to start working on definitions in order to avoid confusions.

WG 7 N 387 offers this definition: “Spectral Efficiency (bps/Hz/cell, bps/Hz/Km2): Throughput provided by unit spectrum resources per cell or unit area.”

The following offers some observations and brief comments regarding this definition:

N 387 definition of Spectral Efficiency is different from the definition used in MCS innovation programs.

More analysis on Spectral Efficiency terminology is needed.

Energy Efficiency is also a key vocabulary frequently used in MCS Innovation. WG 7 N 387 offers this definition: “Energy efficiency (bit/J): The number of bits that can be transmitted per joule of energy.”

The following offers some observations and brief comments regarding this definition:

N 387 definition of Energy Efficiency is different from the definition used in MCS innovation programs.

More analysis on Energy Efficiency terminology is needed.

WG 7 N 387 offers a definition for Cost efficiency:

“Cost-efficiency (bit/dollar): Number of bits transmitted per Unit cost. Compared with 4G system, 5G system needs to be significantly improved in Spectral efficiency, energy efficiency and cost efficiency. Spectral efficiency needs to be increased by 5~15 times; Energy efficiency has increased by over a hundred times. Cost efficiency has increased by over a hundred times.”

Above discussions have revealed discrepancies about some key terms such as Spectral Efficiencies and Energy Efficiencies. These discrepancies will cause confusions and misunderstanding if they are not effectively resolved.

Since WG 7 N 387 definitions are based on well established Mobile communication standards and in special application scenarios, AG 4 does not need to request changes from other organizations. What AG 4 can and should do is to establish its own system of terminologies with clearly defined definitions for terms used in MCS Innovation standard systems. For start, AG 4 can work out definitions for:

MCS

...