Internet-Draft | Service Flow Mapping | February 2025 |
Zhu | Expires 21 August 2025 | [Page] |
This Internet-Draft specifies a comprehensive framework for mapping service flow characteristics to network modal resources in multi-modal intelligent computing networks. It introduces the use of the ALTO protocol for collecting service flow data and leverages an SDN architecture to separate control and data planes. The ALTO protocol facilitates the acquisition of diverse network state information, including data from several SDN domains and dynamic network environments, directly from controllers while keeping the provider's internal details confidential. It then transmits the controller's decisions using a proven method. The document details methods for characteristic identification, intelligent mapping, and continuous optimization, enabling dynamic resource allocation and improved network performance. The framework is designed to support scalable, efficient, and secure operations in environments with complex network loads and diverse service requirements.¶
This Internet-Draft is submitted in full conformance with the provisions of BCP 78 and BCP 79.¶
Internet-Drafts are working documents of the Internet Engineering Task Force (IETF). Note that other groups may also distribute working documents as Internet-Drafts. The list of current Internet-Drafts is at https://datatracker.ietf.org/drafts/current/.¶
Internet-Drafts are draft documents valid for a maximum of six months and may be updated, replaced, or obsoleted by other documents at any time. It is inappropriate to use Internet-Drafts as reference material or to cite them other than as "work in progress."¶
This Internet-Draft will expire on 21 August 2025.¶
Copyright (c) 2025 IETF Trust and the persons identified as the document authors. All rights reserved.¶
This document is subject to BCP 78 and the IETF Trust's Legal Provisions Relating to IETF Documents (https://trustee.ietf.org/license-info) in effect on the date of publication of this document. Please review these documents carefully, as they describe your rights and restrictions with respect to this document. Code Components extracted from this document must include Revised BSD License text as described in Section 4.e of the Trust Legal Provisions and are provided without warranty as described in the Revised BSD License.¶
This Internet-Draft is submitted in full conformance with the provisions of BCP 78 and BCP 79. Internet-Drafts are working documents of the Internet Engineering Task Force (IETF). Note that other groups may also distribute working documents as Internet-Drafts. The list of current Internet-Drafts is available at https://datatracker.ietf.org/drafts/current/. Internet-Drafts are draft documents valid for a maximum of six months and may be updated, replaced, or obsoleted by other documents at any time. It is inappropriate to use Internet-Drafts as reference material or to cite them other than as "work in progress." This Internet-Draft will expire on 17 August 2025.¶
Copyright (c) 2025 IETF Trust and the persons identified as the document authors. All rights reserved. This document is subject to BCP 78 and the IETF Trust's Legal Provisions Relating to IETF Documents (https://trustee.ietf.org/license-info) in effect on the date of publication of this document. Please review these documents carefully, as they describe your rights and restrictions with respect to this document. Code Components extracted from this document must include Revised BSD License text as described in Section 4.e of the Trust Legal Provisions and are provided without warranty as described in the Revised BSD License.¶
This standard aims to provide a comprehensive and systematic specification for mapping service flow characteristics to network modal resources in multi-modal intelligent computing networks. By introducing the ALTO protocol to collect service flow characteristic data and adopting an SDN architecture that separates the control plane from the data plane, this standard supports the creation of stable and efficient mapping templates between application service flows and modal resources.¶
This standard applies to designers, developers, and operators of multi-modal intelligent computing networks, particularly those requiring handling of complex network loads, computing resource demands, and data transmission efficiency in vertical industries. It defines methods for extracting critical service flow characteristics from applications and achieving effective mapping to network modal resources based on these characteristics.¶
Intelligent Computing: AI-oriented computing capabilities for training and executing AI models.¶
Multi-Modal Intelligent Computing Network: A network integrating multiple modalities to serve diverse application scenarios, with computing as the core and network as the foundation.¶
Service Flow: A continuous data transmission process generated by an application or service, including unidirectional (e.g., client-server requests) or multi-directional interactions (e.g., video conferencing).¶
Service Flow Characteristics: Metrics describing application behavior, including throughput, latency, packet loss rate, CPU/GPU utilization, and storage capacity usage across three dimensions: storage, network forwarding, and computing.¶
Elastic Perception Feature Vector: A scalable vector representation dynamically adjusting granularity to characterize multi-dimensional service flow characteristics for flexible resource allocation.¶
Network Modality: A specific network type or configuration optimized for functions such as high bandwidth, low latency, or concurrency.¶
Modal Resource: Basic units constituting multi-modal networks (e.g., computing nodes, switching devices).¶
Feature-Modal Mapping Mechanism: A technical framework for matching service flow characteristics with optimal modal resource combinations.¶
SDN: Software-Defined Networking¶
ALTO: Application-Layer Traffic Optimization Protocol¶
This standard addresses challenges in multi-modal intelligent computing networks, including dynamic workloads, heterogeneous service requirements, and frequent resource state changes. Key objectives:¶
Characteristic Identification: Use AI algorithms (e.g., graph matching, reinforcement learning) to analyze service flow characteristics.¶
Intelligent Mapping: Build an SDN/ALTO-based framework for dynamic resource allocation.¶
Continuous Optimization: Implement feedback loops to refine configurations based on real-time monitoring.¶
The architecture comprises four layers:¶
+------------------------+ +-------------------------+ | Data Collection Layer | | Infrastructure | | +---------------+ | (Polling/Event | +-------------------+ | | | ALTO Server | | Triggering) | | Computing Network | | | +---------------+ |<-Monitoring data->| | Resource Nodes | | | ( Network topology, | | +-------------------+ | | traffic distribution, | | ( Monitoring/ | | link delay/bandwidth, | | configuring resources ) | | resource utilization ) | +-------------------------+ +------------------------+ ^ ^ | | | +-Service Flow Characteristics-+ +--Configuring policies--+ | | v v +-----------------Control Plane ( ALTO Client )-----------------+ | +---------------------------+ +---------------------------+ | | | Data Processing Layer | | Analysis & Decision Layer | | | | +-----------------------+ | | +------------------+ | | | | | Flow Rules generation | | | | Analytical Model | | | | | +-----------------------+ | | +------------------+ | | | | | Link management | | | | Decision Module | | | | | +-----------------------+ | | +------------------+ | | | +---------------------------+ +---------------------------+ | +---------------------------------------------------------------+ ^ | +--Flow rules--+ | v +-------------Data Plane-----------+ | +------------------------------+ | | | Infrastructure | | | | +------------+ | | | | | SDN switch | | | | | +------------+ | | | | ( Forwarding flow rules | | | | and obtain network status ) | | | +------------------------------+ | +----------------------------------+¶
Provides hardware resources (computing nodes, switches) to support feature extraction and configuration.¶
Collects real-time metrics: network topology, traffic distribution, link latency, CPU/GPU utilization, and storage I/O. Supports polling/event-driven mechanisms via SNMP, NetFlow, etc.¶
Constructs service feature topology graphs using adjacency matrices. Nodes represent computing/storage metrics; edges represent network forwarding metrics. Employs distributed stream processing and graph databases.¶
Performs deep analysis using graph neural networks and reinforcement learning to identify optimization strategies (e.g., topology adjustments, load balancing).¶
Defined as three vectors:¶
A three-dimensional tensor: {Service Capability, Controllable Resources, Operational Logic}.¶
The mapping workflow consists of:¶
Feature Extraction: Collect real-time metrics across storage, compute, and network dimensions.¶
Topology Construction: Generate feature graphs with node/edge attributes.¶
Modal Matching: Align service features with modal resources using graph-matching algorithms.¶
Optimization: Adjust configurations (e.g., path rerouting, load migration).¶
Feedback: Continuously monitor performance and update mapping rules.¶
Technical Requirements:¶
The transmission control model employed in this document relies on the default security mechanisms provided by SDN and ALTO protocols. This draft does not alter the default encryption and authentication models as specified in [RFC7149], [RFC7285], [RFC7286] and [RFC7971]. Therefore, the overall security of the service flow mapping system depends on the secure configuration and proper deployment of these underlying protocols.¶
This memo includes no request to IANA.¶
Huanxing Zhu Huazhong University of Science and Technology Wuhan, China Email: huanxingzhu@hust.edu.cn¶
[REPLACE/DELETE] Optional acknowledgements go here.¶
[REPLACE/DELETE] Optional contributors go here.¶