032 RFQ attachment - SOW.docx

Performance Work Statement

Distributed Optimization and Learning Algorithm Design for Next Generation Networks

I. BACKGROUND

The Communications Technology Laboratory (CTL) at the National Institute of Standards and Technology (NIST) promotes the development and deployment of advanced communications technologies through the dissemination of high-quality measurements, data, and research supporting U.S. innovation, industrial competitiveness, and public safety. CTL is currently developing new networking and measurement technologies for next-generation networks, including 5G/6G networks and quantum networks.

OBJECTIVE/PURPOSE: The purpose of this requirement is to obtain contractor support services in developing several new networking and measurement technologies for next-generation networks (e.g., 5G and 6G, and quantum networks) in several key areas. The main objective of this Statement of Work (SOW) is to provide research support and design distributed optimization and learning algorithms for the following:

A. Efficient entanglement routing in quantum networks in the presence of untrusted nodes, B. Optimal resource allocation for End-to-End Quality-of-Service (E2E QoS) and resilience assurance in autonomous networks, focusing on 5G/6G networks, C. Collaborative learning among distributed network nodes under privacy and security constraints; this includes both securing distributed learning and distributed learning for network security.

Additionally, the project also aims at prototyping and deploying these algorithms in suitable testbeds.

The Contractor shall work to complete the development of these algorithms. The Contractor shall work closely with a team of CTL scientists who are working on 6G core networks, where this project is located.

II. SPECIFIC TASKS

The Contractor shall be responsible for project oversight, equipment, administration, and technical execution of this contract. The Contractor shall perform the following specific tasks:

A. Research on entanglement routing models for quantum networks - this includes the following:

1. Identify fundamental differences between routing models in quantum networks and classical networks.

2. Analyze limitations of current entanglement routing models, algorithms, and metrics.

3. Develop a new distributed algorithm for entanglement routing so as to enable long-distance entanglements between concurrent source-destination pairs via untrusted repeater nodes.

4. Provide theoretical analyses and numerical evaluations of the algorithm.

B. Design efficient resource allocation approaches for ensuring End-to-End (E2E) Quality-of-Service and resilience in autonomous networks – this includes the following:

1. Identify limitations in current autonomous networks such as 5G

2. Develop novel distributed resource allocation algorithms that allow network nodes to (i) adapt to network’s dynamic environment, (ii) cooperate to meet E2E service quality goals while managing their local resources autonomously, and (iii) collaboratively respond to epidemic spread of malware/viruses/failures. The design shall aim to support more stringent resilience and E2E requirements in next-G networks.

3. Provide theoretical analyses and numerical evaluations of the algorithms.

C. Research on collaborative distributed learning under privacy/security constraints – this includes the following:

1. Design a new Federated Learning (FL) architecture that enables edge nodes, without sharing their local data, to collaboratively learn a global machine learning (ML) model effectively even in the presence of significant heterogeneity of their computation/communication capabilities and data distributions.

2. Provide theoretical analyses and numerical evaluations of the architecture.

3. Specialize this architecture for network security applications, including distributed network anomaly detection.

4. Analyze metrics and evaluation methods for testing the performance of edge leaning algorithms.

D. Prototype and implement the algorithms from tasks A, B, and C in suitable testbeds, including ANET.

III. DELIVERABLES AND DELIVERABLE DUE DATES

The following deliverables shall be completed during the period of performance. All deliverables are subject to clause 52.227-17, Rights in Data—Special Works.

· Final report documenting algorithms developed, prototype code, and experimentation results

· Progress reports: Every two weeks

· Final report: August 31, 2023

· Final report documenting algorithms developed, prototype code, and analysis results

· Progress reports: Every two weeks

· Final report: August 31, 2023

· Final report documenting algorithms developed, prototype code, and analysis results

· Progress reports: Every two weeks

· Final report: August 31, 2023

· Demonstration and testing on ANET testbed. Test suite and deployment plan (Word document)

· Final report documenting prototype implementations, source code, and experiment setups, and experimentation results

· Progress reports: Every two weeks

· Documentation of testing and test development plans: November 30 2022.

· Final report: August 31, 2023

Standards for Acceptance of Deliverables: The NIST Technical Point of Contact (TPOC) shall provide comments on each deliverable within 14 calendars days from receipt of a given deliverable. The Contractor shall make any needed changes to the deliverables within 14 calendar days from receipt of electronic or written comments from the TPOC. The TPOC will provide written notification when a deliverable is accepted.

IV. PERIOD OF PERFORMANCE

The period of performance shall be for 12 months from September 1, 2022 – August 31, 2023.

V. PLACE OF PERFORMANCE

All the tasks specified in Section III shall be accomplished at NIST, Gaithersburg, Maryland. The Contractor shall work during normal duty hours from Monday through Friday, 9:00 a.m. to 6:30 p.m. except for Federal holidays. Offsite work will also be permitted with advance NIST approval if the Government determines it is advantageous to the Government for progress or performance of the work to take place at an approved telework location.

The Contractor shall visit the campus for badging and to pick up any necessary equipment for telework if applicable. NIST will coordinate any such arrangements with the Contractor. While at the NIST campus, Contractor personnel shall follow NIST Visitor and Contractor Protocols for Mitigating COVID-19 Exposure on NIST Campuses as identified at this link: https://www.nist.gov/about-nist/visit/nist-visitor-and-contractor-protocols-mitigating-covid-19-exposure-nist-campuses

It is the Contractor’s responsibility to be familiar with this site and any updates made to it during the course of the requirement. If Contractor personnel are unable to meet all requirements for campus access identified at the link, work will need to be rescheduled and shall be rescheduled at the Contractor’s expense with consideration to the Government for any delays suffered.

VI. GOVERNMENT FURNISHED PROPERTY, DATA AND/OR INFORMATION

All property, data and information provided by the Government in the performance of this requirement remains the property of the Government and shall be surrendered to the Government upon completion or termination of this requirement. Likewise, all deliverables remain the property of the Government.

Upon award of this requirement, NIST will provide on-site office space, workstation, PC, email and Internet access, telephone, routine office supplies, and access to all facilities needed for the project. The Contractor shall only access low risk general-purpose IT equipment and common network services available to NIST associates.

VII. TRAVEL AND OTHER DIRECT COSTS

Travel and other direct costs are anticipated in relation to this requirement. Cost for travel expense throughout the period of performance shall not-to-exceed a limit of $7,000.00. NIST will provide at least two weeks of advance notice for any travel or other direct cost requirements. All travel and other direct costs shall be approved in advance by the TPOC and will be reimbursed following the guidelines of GSA’s Federal Travel Regulations (FTR) and the Department of Commerce Travel Handbook.

VIII. CONTRACT TYPE

The contract type will be a firm fixed price purchase order utilizing the following CLIN structure:

CLIN 0001: Fixed-Price Tasks 01-04 CLIN 0002: Reimbursable not-to-exceed – Travel and Other Direct Costs $7,000

IX. RISK ASSESSMENT

The HSPD-12 Security Risk Level assigned to this contract is low risk.

X. PERFORMANCE REQUIREMENT SURVELLIANCE MATRIX

- Design of distributed algorithms for entanglement routing

- Design of distributed resource allocation algorithms for E2E QoS and resilience assurance in autonomous networks

- Design of a new Federated Learning architecture

- Design of the overall algorithms and their components.

- Distribution of the design document to the team.

- Presentation of algorithms and their analyses.

- NIST’s GitHub repository of the simulation codes organized in modules with documentation The design document includes an overall architecture figure, entity descriptions composing every component and interactions diagrams between the different components and the data structures used in every component for these interactions.

The code follows standard coding styles with the description of each module at the top and a description of each function right before its code.

The TPOC shall review the progress of the design document, any written report, and codes, and determine its acceptability.

- Final report covering the algorithms, architecture together with its components, and the tests conducted during this project, including:

1. Implementation of algorithms in testbed.

2. Demonstrations and suggestions on further extensions of the designs and implementations.

- Description of algorithms, their implementation, and the main characteristics of their design.

- Presentation of implementations and conclusions that can be drawn based on experiments performed on one or more test deployments.

XI. MINIMUM CONTRACTOR QUALIFICATIONS

It is anticipated that this work will require one (1) Key Personnel as a computer scientist. This labor category title is not mandatory, it is simply a sample labor category that can cover all of the minimum qualification requirements. The proposed key personnel shall have the following minimum education qualifications and experience requirements:

Computer Scientist:

1. Ph.D in computer science, computer engineering, electrical engineering, mathematics, or statistics.

2. Five (5) years of post-Ph.D. experience in convex optimization, machine learning, algorithm design and analysis, and SIS/SIR epidemic modeling.

3. 5 years of experience working with software version control systems such as Git and GitHub.

4. 5 years of experience in independent design, efficient implementation, and testing of all aspects of an optimization or machine learning algorithm in a testbed.

5. Knowledge of 5G networks and mechanisms for supporting quality of services over 5G systems.