Chapter 26 — Network and communications security: architecture alternatives
Lesson map
Lesson 26 published. Map, example, EX076.
flowchart TD
N0["Learning outcomes"]
N1["26.1 Alternatives"]
N2["26.2 Criteria"]
N3["26.3 Evaluation"]
N4["26.4 Justification"]
N5["Sensitivity of the choice"]
N6["Future migration paths"]
N7["Synthesis"]
N0 --> N1 --> N2 --> N3 --> N4 --> N5 --> N6 --> N7
N7 --> EX["Worked example"]
EX --> Q["Exercise EX076"]
Logical drawing
flowchart LR
N0["Learning outcomes"]
N1["26.1 Alternatives"]
N2["26.2 Criteria"]
N3["26.3 Evaluation"]
N4["26.4 Justification"]
N5["Sensitivity of the choice"]
N6["Future migration paths"]
N7["Synthesis"]
N0 --> N1 --> N2 --> N3 --> N4 --> N5 --> N6 --> N7 --> Q["EX076"]
[Learning outcomes]
|
v
[26.1 Alternatives]
|
v
[26.2 Criteria]
|
v
[26.3 Evaluation]
|
v
[26.4 Justification]
|
v
[Sensitivity of the choice]
|
v
[Future migration paths]
|
v
[Synthesis]
|
v
[Worked example] --> [Exercise EX076]
Learning outcomes
You will (1) compare alternative architectures; (2) evaluate against criteria; (3) justify a choice.
26.1 Alternatives
Kofi compares three options: A) central firewalls at the main office with branches routing all inter-zone traffic there; B) firewalls at every office; C) a cloud-delivered security service with branch connectors.
For each major decision, engineers consider several realistic alternatives, such as different zone counts or VPN approaches, so the final choice is a comparison rather than a default.
Including a do-minimum option, such as only isolating guest Wi-Fi, gives a useful baseline: every more ambitious alternative must show enough extra benefit to justify its extra cost and disruption. Considering at least three options also guards against the common habit of comparing a preferred design with an obviously poor one.
26.2 Criteria
Criteria: security, availability, cost, latency and operational effort.
Criteria are the qualities used to compare alternatives: security benefit, cost, operational effort, disruption during rollout, and fit with existing skills and equipment. Agreeing criteria and their weights with the council before scoring prevents the analysis being shaped to favour a preferred option. Each criterion needs a clear scale, such as low, medium or high, with a short description of what each level means, so scores are consistent.
Weights should reflect the council's priorities rather than the engineer's.
26.3 Evaluation
Option A is cheapest but makes the main office a single dependency and adds latency for branch-local traffic. Option C reduces hardware but depends on internet links and a new skill set. Option B costs more but keeps local traffic local and survives WAN outages. Kofi chooses B with small firewalls at branches.
Scores are recorded with brief evidence for each.
Each alternative is scored against every criterion, with the evidence noted, such as lab measurements, vendor documentation or operator interviews. Weighted totals give a ranking, but Kofi also checks how sensitive the ranking is to the weights: if a small change in one weight reverses the order, the decision is close and deserves discussion rather than being presented as clear-cut.
26.4 Justification
The choice is recorded in a decision record with the evaluation table, so future engineers understand why.
The justification explains why the chosen alternative is best against the agreed criteria, acknowledges its weaknesses, and notes what evidence would change the decision. Writing it plainly lets council managers understand and approve the choice. It also helps future engineers: if conditions change, such as a large budget increase, they can see whether the original reasoning still holds and revisit the decision with confidence.
Sensitivity of the choice
Kofi tests whether the choice of option B depends on the weights. If cost is weighted twice as heavily, option A edges ahead; if availability is weighted more, B's lead grows. He presents this sensitivity to the council's managers, explaining that the choice depends on how much they value branch independence during WAN outages. They confirm that keeping branch services running matters more than the extra cost, which makes the decision theirs, informed by the analysis, rather than his. The sensitivity table is kept in the decision record so the choice can be revisited if priorities change.
Future migration paths
The chosen architecture should not block future change. Kofi notes that branch firewalls can later forward traffic to a cloud-delivered service if the council adopts one, and that the zone model and policy would carry across. Recording this migration path reassures managers that option B is not a dead end and shows how today's design relates to longer-term zero trust ambitions described in NIST SP 800-207. He also notes which parts would need replacing, such as on-site inspection features, so future costs are visible early.
Worked example
The table scores B highest on availability and latency, second on cost.
Synthetic case. Scores are synthetic.
Errors and limits
Scoring without weights hides priorities; agree weights with stakeholders.
Exercises
- EX076 — Option A weakness. What is the main weakness of central firewalls only? Rubric: 3 = single dependency and latency;; 2 = dependency;; 1 = vague;; 0 = none.
- EX077 — Option C risk. What new dependency does option C introduce? Rubric: 3 = internet links and new skills;; 2 = cloud;; 1 = vague;; 0 = none.
- EX078 — Weights. Why agree criterion weights with stakeholders? Rubric: 3 = reflect priorities;; 2 = fairness;; 1 = vague;; 0 = none.
Synthesis
Comparing alternatives systematically makes the chosen architecture defensible.
Continue studying — chapter 26 of 50
← Chapter 25 (draft) · Course chapter index · Chapter 27 (draft) →