Network Function Virtualization (NFV) enables service providers to deploy network functions as software on general-purpose infrastructure, increasing flexibility and resource efficiency. Many of these services require an ordered sequence of virtual functions, known as a Service Function Chain (SFC), to be embedded and routed across the substrate under tight resource constraints. The Service Function Chain Embedding (SFC-E) involves mapping virtual sequences of network services and their virtual links onto a physical substrate network. Many solutions have been developed for this problem, ranging from exact optimization methods to heuristic and evolutionary algorithms. More recent work has focused on deep reinforcement learning (DRL) to model the problem as a sequential decision process. However, standard DRL requires extensive sampling and faces sparse, delayed rewards, obscuring which decisions caused acceptance versus rejection. In this work, we propose Alpha-SFC, an AlphaZero-style framework that combines a graph-based policy-value transformer with a Monte Carlo Tree Search (MCTS) planner. Our contributions are twofold. First, the neural network guides the search, enabling the system to focus on promising placements and prune early with a clearer evaluation than pure MCTS alone. Second, we integrate in-tree multi-path routing during learning: at each node placement, the planner enumerates up to k loopless shortest candidates for each newly realized virtual edge and commits during expansion to the first bandwidth-feasible path, allowing search and reservation to co-evolve inside MCTS. This coupling—rather than performing post-hoc routing outside the planner—consistently improves acceptance and overall embedding efficiency across a range of routing budgets.

Our recent teaching of programming courses has witnessed significant non-success rate among freshmen, echoing the observations reported by other institutions. This work represents a systematic study on the methodologies to intervene with at-risk CS/IT freshmen in computer programming, which integrates best-practices in the literature and our innovations. Grounded on Self-Determination Theory, we propose methodologies that embrace four major principles. First, embed fundamental concepts in a CS0 class. Second, adopt themes in selfmotivated topics that reflect intrinsic interests of freshmen. Third, progression from low-floor block languages to general languages. Fourth, adopt hybrid assessment tools that encourage handson, collaboration and ipsative improvement. A CS0 course is implemented to validate the soundness of the proposed methodologies. Based on student questionnaire feedback and hypothesis testing, this work uncovers three major findings in intervening at-risk freshmen in programming. First, core programming concepts conveyed in block languages can facilitate the learning of respective concepts in general languages. Second, self-motivated topics contribute to students’ learning of programming, and can retain their interests post-course. Third, ipsative assessment can considerably contribute to students’ mastery of programming concepts.

The last several years have witnessed a blossom of generative AI technologies with wide applications from business to healthcare. As revealed recently, generative AI tools can be incorporated in cybersecurity practices as both offensive and defensive vectors. This paper represents an initiative to empower Ethical Hacking (EH) education with ChatGPT-infused curriculum. As the first attempt of this nature, the proposed curriculum design has a few major innovative features. First, our design identifies and incorporates four ChatGPT-inspired EH areas including ChatGPT-powered Attack, ChatGPT-powered Defense, ChatGPT-powered Security Tasking, and ChatGPT-rooted Security. Second, our design is aligned with major EH certifications to prepare students to be industry ready. Third, we tackle both physical constraints (e.g., budget and space) and technical constraints (e.g., teaching offensive skills in compliance with campus policies), resulting in a cost-effective and IT-compliant design. Fourth, our design identifies and presents a course flow that integrates the diverse topics and ad hoc techniques for EH. Additionally, we discuss students’ performance and feedback on the design, and share learned lessons, with the hope of enlightening similar endeavors of other institutions.

The 2023 CS curriculum by ACM, IEEE, and AAAI identifies security as an independent knowledge area that develops the “security mindset” so that students are ready for the “continual changes” in computing. Likewise, the curriculum emphasises the coverage of “uses”, and “shortcomings/pitfalls” of practical AI-tools like ChatGPT. This paper presents our endeavors to approach those goals with the design of an Information Security course. Our course design bears the following distinct features: Certificate-readiness, where we align the knowledge areas with major security/ethical hacking certificates; Coverage of ChatGPT, where the uses of ChatGPT for assisting security tasks and security issues caused by ChatGPT usage are both addressed for the first time in the teaching; “Learn defending from attackers' perspective”, where labs of both offensive and defensive natures are developed to equally sharpen ethical hacking and hardening skills, and to facilitate the discussion on legal/ethical implications; Current and Representative, where ajust-enough set of representative and/or current security topics are selected in order and covered in respective modules in the most current form. In addition, we generalize our design principles and strategies, with the hope to shed lights on similar efforts in other institutions.

There have been increasing efforts in recent years to apply visualization tools for teaching CS/IT courses. Particularly, the abstractness of programming concepts can be mitigated with visualization tools by providing a WYSIWYG (“what you see is what you get") experience. Visualization tools for programming can provide diverse coverage ranging from basic programming concepts such as variables to advanced concepts in data struc-tures and algorithms. Given the varieties of the tools and the fast software updates, we are motivated to identify visualization tool(s) that are current, supportive (e.g., in both breadth and depth of coverage), and friendly (e.g., no learning curve) to facilitate the teaching of programming. Motivated by this, this paper provides a systematic overview of visualization tools according to four defined metrics. Moreover, we share our best-practices in adoption: a hybrid approach for selecting a combination of visualization tools in an introductory programming course and a data structures course. In addition, we discuss students' feedback, our strategies and lessons learned in adopting visualization for teaching programming.

In Network Function Virtualization (NFV), traditional purpose-built proprietary middle-boxes are replaced with virtual network function (VNF) instances that are deployed over generic hardware; network traffic flows are steered through a sequence of VNF instances (i.e., that forms a service function chain) after resolving the VNF Forward Graph Embedding (FGE) problem. NFV eases the network management and fosters the agile deployment of network services, thus enabling a foremost solution to address issues of traditional enterprise networks. With the everlasting Internet traffic growth, in this work, we extend the provision of NFV upon Elastic Optical Networks (EONs) that deliver abundant elastic and fine-granular bandwidth for connections between VNFs. We systematically examine the integrated NFV-EON architecture, and investigate one key enabler, namely the FGE-EON problem, where one has to coordinate the allocation of computing resources (as in a typical FGE problem) and the assignment of optical bandwidth (i.e., spectrum allocation in EONs). The contributions of this work are multi-fold. The FGE-EON problem is proven to be NP-Complete for the first time. An Integer Linear Programming (ILP) model is developed to address FGE-EON problem of small instances. Despite its complexity, with novel modeling techniques, the proposed ILP model maintains a compact structure even after accounting for VNF sharing and VNF co-location together for the first time. For large scale problems, we design a heuristic algorithm based on layered graph that is shown to be near-optimal in our evaluations.

In recent years, many universities have made it a priority to provide students with more opportunities for undergraduate research. This growth has been particularly notable in STEM, with undergraduate research appearing in a variety of forms (e.g., as part of a course, as a multi-day event, or as a summer-focused project). Some endeavors, like the National Science Foundation’s Research Experiences for Undergraduates (REU) program, are large-scale ventures, while others are localized. Regardless of scope, the benefits of adopting the recognized high-impact practice of undergraduate research are multifold. First, students who are exposed to state-of-the-art research can acquire research and professional skills; increase their knowledge, interest, and confidence in the subject; and improve their academic performance. Second, undergraduate research fosters the professional development of faculty advisors by providing a stage for sharpening skills in research guidance and advising, as well as working collaboratively with students to advance their work. Third, it benefits the field of research at large by increasing student retention and persistence, promoting recruitment for graduate programs, and diversifying the demographics of the field. These benefits are well evidenced in the literature. A successful undergraduate research experience, however, relies on a design that accounts for the particular constraints of the host institution. Like many other non-R1 institutions, we are limited in terms of budget, facilities, space, and faculty and student time to dedicate to undergraduate research projects, and we also face the uncertainty of short-term research (due to its experimental nature). In this work, we present our design for a summer-focused undergraduate research (SUR) program in CS/IT that has been implemented on a regular basis over the past decade. The program aims to provide the key features of a well-rounded graduate research experience, scaled down to fit the constraints under which we are operating. Additionally, it serves as an opportunity to increase participation in CS/IT by students from under-represented minority groups and to facilitate collaboration with faculty and students in other STEM fields. In addition to presenting the design, we discuss the outcomes, summarize the design strategies, and share lessons learned from the program, with the hope of motivating the development of similar endeavors at other institutions.

The abstract nature of programming concepts in CS/IT courses causes challenges for undergraduate students to grasp them. Visualization tools can mitigate this challenge by providing a ”what you see is what you get” experience. Given the diversity of the tools and the fast pace of software updates, it remains a question to select the best-fit tools that can be of current use for the respective programming concepts. This study aims at providing an assessment and classification of visualization tools based on four defined metrics which suggests a hybrid approach for selecting combination of visualization tools in teaching.

The recent decade has witnessed a tremendous growth of Internet traffic, which is expected to continue climbing for the foreseeable future. As a new paradigm, Spectrum-sliced Elastic Optical Path (SLICE) networks promise abundant (elastic) bandwidth to address the traffic explosion, while bearing other inherent advantages including enhanced signal quality and extended reachability. The fundamental problem in SLICE networks is to route each traffic demand along a lightpath with continuously and consecutively available sub-carriers, which is known as the Routing and Spectrum Allocation (RSA) problem. Given its NP-Hardness, the solutions to the RSA problem can be classified into two categories: optimal solutions using link-based, path-based, and channel-based Integer Linear Programming (ILP) models, which require extensive computational time; and sub-optimal heuristic and meta-heuristic algorithms, which have no guarantee on the solution quality. In this work, inspired by a channel-based ILP model, we propose a novel primal-dual framework to address the RSA problem, which can obtain a near-optimal solution with guaranteed per-instance closeness to the optimal solution.

In the context of virtualization, one critical task that has to be addressed is the mapping of the virtual service request (that consists of virtual nodes and links) to the virtualization-enabled physical network. In this work, we comprehensively study the service mapping problem with the consideration of node re-visitation, which supports the hosting of more than one virtual nodes on the same physical node. Our study fills a gap of the literature by exploring strategies to best utilize node re-visitation for reduced resource consumption and service blocking in the resulting problem of service mapping with node re-visitation (SMNR). Both path-based and link-based Integer Linear Programming SMNR models are designed and evaluated in our simulation.

This chapter identifies and discusses the learning outcomes to be achieved because of hands-on lab exercises using ethical hacking. It discusses the ethical implications associated with including such labs in the information security curriculum. The discussion is informed by analyses of log data on student malicious activities, and the results of student surveys. The examination of student behavior after acquiring hands-on offensive skills shows that there is potentially a high risk of using these skills in an inappropriate and illegal manner. While acknowledging the risk and the ethical problems associated with teaching ethical hacking, it strongly recommends that information security curricula should opt for a teaching approach that offers students both offensive hands-on lab exercises coupled with ethical practices related to the techniques. The authors propose steps to offer a comprehensive information security program while at the same time minimizing the risk of inappropriate student behavior and reducing institutional liability in that respect and increasing the ethical views and practices related to ethical hacking.

We address the challenge of teaching a database course for freshmen CS/IT majors: namely their limited background, coupled with the material's mixture of language (e.g., SQL) and abstract theory (e.g., normalization) - hard even for upperclassmen. In this paper, we propose a new two-thread model: one thread of theory and another of practice which are relatively independent and self-contained (compared to a typical "Lecture + Lab" design). To provide students extensive exercises on SQL, the practice thread has labs that expose them to SQL statements starting Week 1. When applicable, labs purposely anticipate related components in the theory thread. Also we approach abstract concepts (e.g., normalization) using an example-based method that generalizes patterns extracted from examples. We have applied this two-thread model with our freshmen for the last few semesters, and their responses have been very positive. We share some of this feedback as well as the learned lessons with the hope to enlighten the teaching of database courses and other freshmen courses in other institutions.

An Ethical Hacking (EH) course not only is a critical component for a Cybersecurity program but also an essential preparation for CS/IT majors towards career paths as security professionals. We face two major challenges when developing an undergraduate EH course, including the setup and choice of the lab design, and the choice and organization of covered topics for this course. On one hand, we have limited space, budget and technical support for a course that relies heavily on hands-on exercises. Given the nature of this course, the lab activities are often 'offensive' and lab operations demand administrative privileges, which cause compliance issues with the university's IT policies. On the other hand, given the vast variety of topics and the fast pace of the field, it is difficult to select and organize an essential set of knowledge units to ensure that students are exposed to current technologies and prepared to be industry-ready. We adopt two major design principles to address these challenges correspondingly. First, our choice of a hybrid Virtual Machine (VM)-based and Web-based labs provides students the full set of privileges to perform lab activities without posing threats to the campus network. The Web-based labs remove high cost of hardware and avoid overwhelming installations and configurations for the lab. Second, given the diversity of topics and fast developments in this field, we choose topics based on four criteria: representative, current, certification-related, and foundations for other covered concepts. The chosen topics are aligned with three EH certificates, and organized into twelve modules with clear inter-module and intra-module logic. This paper details the curriculum of this EH course and elaborates how our design principles are entailed in the course.

Network virtualization (NV) is considered to be a promising solution to address the ossification of the current Internet infrastructure. With the recent advancements in optical virtualization, network virtualization can be further extended to virtualization-enabled optical substrate networks that supply connectivity-as-a-service. This integration of NV and optical virtualization is referred to as optical-based network virtualization in this work, which warrants a future-proof Internet that is not only technology-friendly (i.e, due to the service and hardware decoupling in NV) but also bandwidth-abundant (i.e., with optical networking). In optical-based network virtualization, it remains to resolve the classic virtual network embedding (VNE) problem (i.e., a NP-Complete problem) with the extra dimension of complexity from physical characteristics of the specific type of optical network (e.g., WDM or SLICE). In this work, we study this variant of the VNE problem, namely Optical Virtual Network Embedding (OVNE), in SLICE-based network virtualization. We avoid addressing the OVNE problem with simple add-on constraints to VNE but address it with two strategies: first, exploring the OVNE problem structure at different granular of network elements (e.g., path-channel and channel graph) to mitigate the complexity; second, designing a path growing process that solves the OVNE model with substantially reduced variables. The proposed approach is evaluated and compared to other representative path-based schemes with demonstrated improvements. The proposed approach can also obtain near-optimal solution to the OVNE problem with guaranteed closeness to the optimal solution.

To deliver an end-to-end service in the Network Function Virtualization (NFV) environment, Service Function Chain (SFC) is used to define the sequence of Virtual Network Functions (VNFs) that needs to be applied to the data stream of the service. In the literature, SFC is typically regarded as a linear topology that represents the sequence of the VNFs that a data stream is navigated through. Our recent study, however, reveals a group of real-life use cases for which a linear topology is in-sufficient to describe the complexity of the SFC: the data stream can branch out into multiple streams due to multiple reasons (e.g., load-balance the web/non-web traffic). In this paper, we study representative use cases with presence of branching, categorize traffic branching into three major categories, and introduce policies to specify the service requirement that involves traffic branching. In addition, we define the resulting problem and discuss directions for further exploration.

Network Function Virtualization (NFV) is an emerging technology that promises to address issues in traditional middleboxes, providing service flexibility and reduced cost. NFV decouples network functions from the proprietary middlebox hardware, thus allowing the network providers to implement network functions on virtual machines running in standard servers. To deliver an end-to-end service in the NFV environment, Service Function Chain (SFC) is used to define the sequence of actions or Virtual Network Functions (VNFs) that needs to be applied to the data stream of a service delivery demand. Typically, the SFC is regarded as a linear topology that consists of a group of VNFs that the data stream is navigated through. However, our recent exploration on a group of real-life use cases reveals that the data stream can branch out to multiple data streams at certain VNFs (e.g., a load balancer). This branching indicates that the packet flow may be directed to multiple different paths, resulting in a more complex mesh-like service function graph. In this paper, we define the resulting problem as Branching-Aware Service Function Graph embedding and routing (B_SFG). We study how to efficiently accommodate user's B_SFG requests in the physical network while considering the constraints of function dependency, branching requirements, computing resources of virtual nodes and bandwidth demand of the virtual links. We propose a novel algorithm, B_SFG_CM, which jointly optimizes the processes of node and link mapping of a request. B_SFG_CM employs the proposed techniques of dependency sorting and layering that effectively take into account the node dependency, branching points and the resource status of the physical network. Our experimental results show that the proposed algorithm significantly outperforms the scheme based on the traditional sequential node/link embedding in which the node and link mapping is done sequentially.

Spectrum-sliced Elastic Optical Path (SLICE) networks are considered as a promising solution for the next-generation optical networking. Compared to traditional WDM networks, SLICE networks not only bear inherent advantages in signal quality and reachability, but also possess elasticity and flexibility in bandwidth management, thanks to the fine spectrum allocation in the unit of sub-carrier. When accommodating traffic demands in SLICE networks, the routing and spectrum allocation (RSA) assigns continuous and consecutive sub-carriers to form a spectrum path (SP), and determines the guard-bands that separates the SP from other SPs sharing any common links. Given the NP-Completeness of the RSA problem, it attracts considerable research efforts in recent years. The existing solutions to the RSA problem can be broadly classified into two categories: optimal solution based on Integer Linear Programming formulations; and heuristic algorithms. The former has the well-known issue of extensive computational time, thus limiting its applicability to network instances of practical interests. The latter can obtain near-optimal solution but can never provide any guarantee on the quality of the solution (i.e., the closeness to the optimal solution). In this paper, we aim to fill this gap by designing a new and novel solution for the RSA problem, with which a near-optimal solution can be found efficiently with guarantee on its closeness to the optimal solution.

Routing and Spectrum Allocation (RSA) is the key problem in Spectrum-Sliced Elastic Optical Path (SLICE) networks. The difficulty of RSA problem lies on three factors: first, the allocated sub-carriers have to be continuously available along each established spectrum path; second, the allocated sub-carriers have to be consecutive in the spectrum domain as implied by the OFDM technology of SLICE networks; and third, sub-carriers of spectrum paths sharing the same fiber have to be separated by the guard-band that is determined at run-time. As a decision problem, the RSA has been proven to be NP-Complete. In this work, we study an optimization version of the RSA problem with the goal of maximizing the revenue from the accommodated requests. We present Integer Liner Programming (ILP) formulations for the problem, namely Routing and Spectrum Allocation with Optimal Revenue (ROR). Also, we present detailed design of a framework that utilizes techniques of relaxation, decomposition and auxiliary graphs, which can be employed to obtain a near optimal solution that has a per-instance guarantee on the closeness to the optimal solution.

The teaching of Information Security has well-known difficulties. On one hand, security issues can arise in a broad spectrum of fields, and it is challenging to chose a subset of topics to fit into one (undergraduate) course as well as maintain a clear flow among the chosen topics. On the other hand, given the rapid advancement in the field of Information Security, it requires a continuous and pro-active effort to keep the course up-to-date. In addition, our prior teachings have also witnessed other challenges when teaching Information Security for undergraduate students. First, students are naturally interested in security skills with an attacking nature. A course design that favors the coverage of defensive skills thus may not be able to strongly stimulate the students' learning. Second, the job market places a large weight on certificates, which, should also be strongly tied to and reflected in the course curriculum of Information Security. To address above issues, we design a new course curriculum for an undergraduate level Information Security course. Our design possesses a few major features and novelties. First, we organize the covered security topics in a security spectrum and strike a balance of theory and practice among the topics. Second, we introduce modules to add build-in natures for actively keeping the covered security topics current. Third, we develop a group of labs that have equal emphasis on offensive and defensive skills to stimulate the critical thinking and interests of the students while addressing the associated legal and ethical implications. Fourth, we incorporate key knowledge set from entry level certification into the curriculum to assist students in obtaining security certificates.