Applied Computer Networking
Network engineering and operations are now largely machine-assisted: telemetry is triaged by learned detectors, configuration and remediation are drafted by AI agents, and the enterprise network has become the delivery and control path for AI services. The course keeps its networking identity and adds the evaluation, validation, and accountability skills a practitioner needs to supervise that automation, including the rollout and rollback discipline a machine-proposed change must clear before production.
Current description → proposed description
This course is an in‐depth view of data communication and networking, ranging from the primitive historical approaches to the ever changing modern state of the field. It includes principles of network design, using a top‐down approach and focusing on technologies used in the Internet. It will help students learn to design network‐aware applications using sockets, threading, and concurrency. It will help students understand how the Internet works, from the transport layer down to the physical layer. It will help students prepare for future positions in research and development by introducing them to the latest research in Internet technologies. It will help students become better writers by emphasizing written work where possible. It will also help students apply networking technology in ways that can enrich their lives and assist in spreading the Gospel.
This course offers a graduate-level, in-depth treatment of data communication and networking, tracing the field from its historical foundations to its ever-changing modern state. Using a top-down approach centered on the technologies of the Internet, students examine principles of network design and the operation of the stack from the transport layer down to the physical layer, and they design network-aware applications using sockets, threading, and concurrency. The course extends that foundation to contemporary practice, in which networks are instrumented continuously and operated with machine assistance. Students work with packet captures, flow records, and streaming telemetry; evaluate machine-learning traffic classification and anomaly detection against base rates, false-positive cost, and drift; and judge AI-generated configuration and remediation against the validation, staged-rollout, and rollback criteria a machine-proposed change must satisfy before it reaches production. Attention is given to the transport and egress path on which enterprise AI services depend, including long-lived token-streaming connections to external model and embedding APIs, and to the privacy, transparency, and accountability questions raised by automated inspection and automated enforcement. Written work is emphasized throughout in design documents and incident reports, preparing students for positions in research and development. Students also consider the stewardship of networked infrastructure as a calling that enriches lives, serves the neighbor, and assists in spreading the Gospel.
What changes
- Graduate register and depth distinguished from the undergraduate networking course
- Telemetry pipelines: packet captures, flow records, and streaming telemetry as working data
- Operational evaluation of ML traffic classification and anomaly detection against drift and false-positive cost
- Transport and egress path engineered for enterprise inference workloads
- Privacy, accountability, and rollback limits on automated inspection, machine-generated configuration, and automated enforcement
6 proposed outcomes, mapped to 7 program outcomes
Each outcome below is written to be observable and assessable, and each is mapped to the program learning outcomes for which it produces evidence.
Students will be able to analyze end-to-end Internet behavior across the transport, network, link, and physical layers by collecting and normalizing packet captures, exported flow records, and active performance measurements from a live path, and reasoning from them to explain how addressing, routing, congestion control, and queueing determine application-level latency, loss, and throughput.
The CLO requires students to collect, normalize, and reconcile measurements from several distinct instrumentation sources - captures, exported flow records, and active probes - and to analyze them to explain observed network behavior, which is the multi-source data acquisition and analysis in service of a real problem that PLO 2.1 names.
Students will be able to implement a concurrent, network-aware service using sockets, threading, and transport-layer protocols, designing its transport path for an enterprise inference workload through long-lived streaming connections that consume a token-by-token model response, connection reuse, retry and rate-limit backoff against the provider's published limits, explicit latency and throughput budgets, and controlled egress to external model and embedding APIs.
The CLO's assessable artifact is a service that actually calls external model and embedding APIs, consumes a streamed token-by-token response, and backs off against the provider's published rate limits inside a stated latency and throughput budget, which is the structured enterprise deployment of LLMs through APIs and embeddings that PLO 6.1 names, earned at the network layer this course owns.
Students will be able to evaluate a deployed machine-learning traffic classifier or anomaly detector by selecting metrics appropriate to a low-base-rate setting, including precision at the operating threshold, alert volume, and analyst cost per false positive, diagnosing degradation caused by traffic drift, pervasive encryption, and adversarial evasion, and specifying the threshold change or retraining plan that follows from that diagnosis.
Selecting metrics fit for a low-base-rate detection problem, diagnosing the model's limitations under drift and evasion, and deriving a threshold or retraining decision from that diagnosis is exactly the performance assessment and iterative, diagnostically driven refinement described in PLO 6.2.
Students will be able to critique a network monitoring, configuration-automation, and enforcement design, including deep packet inspection, encrypted-traffic analysis, machine-generated configuration and remediation, and machine-initiated blocking or quarantine, against the privacy intrusion it imposes, the transparency owed to the people it observes, the distribution of its error costs, the pre-deployment validation, staged-rollout, and rollback criteria a machine-proposed change must satisfy before it reaches production, and the human escalation path required before consequential action.
Judging inspection, agent-drafted configuration and remediation, and machine-initiated enforcement against privacy intrusion, transparency to those observed, who bears the cost of an error, the validation and rollback gate a machine-proposed change must clear before production, and the human escalation that preserves accountability is the ethical and legal integrity evaluation this outcome demands; the workbook states this same outcome twice in word-for-word identical terms, under Christian Faith as PLO 1.1 and under Integrated Disciplinary Knowledge as PLO 3.2.
Students will be able to compose network design documents and post-incident reports that convey telemetry findings, detector limitations, and residual risk accurately to both engineering and executive audiences, stating plainly where AI-assisted analysis or AI-generated configuration contributed, how each was corroborated before deployment, and what remains uncertain.
Writing the same findings, the limits of the AI-derived evidence and AI-generated configuration behind them, and their risk and ethical implications for an engineering audience and again for a non-technical executive audience is the clear and responsible communication of AI results and ethical considerations to both audiences that PLO 5.1 specifies.
Students will be able to justify decisions about network access, capacity, cost, and reliability by reference to who is able to reach the networked services those decisions carry, AI-dependent services among them, framing the stewardship of shared networked infrastructure as a Christian vocation exercised in service of the neighbor and the mission of the church.
Arguing for network access, capacity, cost, and reliability decisions on the grounds of who gains or loses reach to the services they carry, AI-dependent services included, and naming infrastructure stewardship as a calling that serves the neighbor and the church's mission, is the vocational articulation guided by a Christian worldview that PLO 1.2 requires, and it carries forward the course's existing catalog commitment to networking that enriches lives and assists in spreading the Gospel.
Program outcomes this course reaches
Filled cells are program learning outcomes with at least one supporting course learning outcome in this course. Sparse coverage is expected — no single course carries all twelve.