27 WSN Coverage: Fundamentals
WSN coverage fundamentals, coverage planning, area coverage, point coverage, barrier coverage, WSN coverage evidence
27.1 Start With the Field Story
Coverage starts with the question “covered for what?” A field can be covered for area monitoring, point events, barriers, paths, or redundancy, and each meaning needs a different sensing model and different proof.
27.2 In 60 Seconds
WSN coverage fundamentals turn a monitoring goal into a reviewable coverage claim. A good claim says what must be sensed, what kind of coverage is needed, which sensing model is being trusted, what redundancy is required, which sensors are active, how readings reach the sink, and when the decision must be retested.
This chapter is the planning gateway for the coverage sequence. It avoids treating coverage as a single formula. The right answer depends on the application, environment, active state, and evidence standard.
27.3 Learning Objectives
By the end of this chapter, you will be able to:
- Translate a monitoring goal into an area, point, barrier, or path coverage claim.
- Choose a sensing model that matches the environment and evidence available.
- Explain how redundancy, duty cycling, and connectivity affect coverage acceptance.
- Identify common ways a coverage plan drifts between design and deployment.
- Build a compact planning record that links to deeper coverage analysis.
27.4 Quick Check: WSN Coverage: Fundamentals
27.5 Coverage Planning Questions
Start coverage planning with questions, not with node counts.
What must be sensed? Name the event, target, region, route, boundary, or condition that matters to the learner or operator.
Where must it be sensed? Distinguish continuous regions, discrete assets, perimeter crossings, routes, and priority zones.
When must it be sensed? State whether the claim applies continuously, periodically, during alarms, during maintenance, or only during a mission window.
How much confidence is needed? Choose the coverage level and evidence standard from consequence, not from habit.
Which state is reviewed? All-awake maps, active sets, rotation schedules, degraded states, and mobile states can produce different answers.
How does data leave? A sensed event is not useful until an active node can deliver the reading to a sink, gateway, or local decision point.
27.6 Planning Route
Figure 1 shows the order of decisions. A team should not jump from the monitoring goal directly to a sensor count.
27.7 Coverage Types
Coverage type controls the rest of the review. A design that is excellent for one type can be wrong for another.
Area coverage Reviews whether a continuous region has acceptable sensing evidence. It is common in environmental, agricultural, and process monitoring.
Point coverage Reviews whether named targets, assets, entrances, valves, machines, or measurement points are covered.
Barrier coverage Reviews whether a crossing path can pass through a boundary without enough detection evidence.
Path or route coverage Reviews whether a moving asset, inspection path, pipeline, corridor, or route segment has enough sensing support.
The coverage type should appear in the first line of a review record. If it is hidden, the team may accidentally optimize the wrong problem.
27.8 Sensing Model and Evidence
The sensing model explains what “covered” means. The evidence explains why the team believes that model applies.
Boolean model Useful for simple first-pass maps. It treats sensing as acceptable inside the stated condition and unacceptable outside it.
Probabilistic model Useful when detection fades with distance, obstruction, orientation, weather, interference, calibration, or target strength.
Measured evidence Field readings, calibration checks, site tests, or inspection passes can override a convenient planning assumption.
Simulation evidence Simulations can compare options, but the record should name the assumptions and what still requires field confirmation.
The model can change over time. A greenhouse fills with foliage, a warehouse adds shelving, a gateway moves, or a perimeter grows vegetation. A fundamentals review must name what change reopens the decision.
27.9 Redundancy and k-Coverage
Redundancy should match consequence and failure mode.
1-coverage One acceptable sensing source may be enough for low-consequence trends or periodic sampling.
k-coverage Multiple acceptable sensing sources support fault tolerance, confidence, and rotation, but only if they are meaningfully independent.
Priority zones Different zones can require different redundancy. Uniform high redundancy can waste energy and budget.
Independence Redundant sensors should not share the same obstruction, mount, relay, gateway, battery domain, or maintenance risk.
k-coverage is not a magic safety label. The record should explain what failures the redundancy tolerates and what failures still break the claim.
27.10 Energy and Operating State
Coverage that only exists when every node is awake is often not the deployed coverage.
All-awake design Useful for the first boundary of possibility, but it often overstates deployed coverage.
Active-set design Reviews the subset of sensors awake during the current operating state.
Rotation design Reviews the weakest scheduled step and the transition between steps.
Degraded design Reviews what remains true after a failed node, blocked sensor, low battery, moved gateway, or missed maintenance window.
Energy planning is part of coverage, not a separate afterthought. Sleep schedules, sampling intervals, local processing, and routing load can all change which coverage claim is true.
27.11 Claim-State Boundary
The merged coverage and core-concepts chapters reinforce a simple rule: coverage is accepted for one stated operating state, not for every state the network might enter.
Keep these boundaries explicit:
- a coverage claim names the region, target, boundary, route, or event class and says when the claim applies
- sensing range is a model assumption that can be Boolean, probabilistic, exposure-style, or field-evidence based
- a coverage hole is any region, target, or route segment outside the accepted sensing condition in the reviewed state
- k-coverage is a redundancy level applied to a coverage type, and the sensors should not all share the same obstruction, relay, power domain, or maintenance risk
- connectivity is separate from sensing coverage: active sensors must be able to deliver evidence to a sink, gateway, or local decision point
- all-awake maps, active sets, rotation schedules, degraded states, and mobile states can each require a separate evidence record
Do not count sleeping, failed, uncalibrated, disabled, unreachable, or out-of-role sensors as live coverage unless the accepted rule states exactly why they still support the claim.
27.12 Evidence Record
Figure 2 shows the minimum record for accepting a coverage plan. The record should be short enough to maintain and explicit enough to retest.
27.13 Worked Review: Storage Facility
A storage facility wants WSN coverage for temperature alarms in three rooms. The initial map shows one sensor per room and a gateway in the office.
Coverage type Area coverage for each room, with priority point coverage near the stored material.
Sensing model Measured evidence is needed because shelves can block air movement and create hot spots.
Operating state The review must show the sensor sampling schedule and gateway path during closed hours.
Decision Accept the basic route, but require a hot-spot walk test and a retest trigger after shelf layout changes.
27.14 Worked Review: Outdoor Perimeter
A site wants alerts when equipment crosses a boundary. The first proposal uses area coverage language, but the real requirement is crossing detection.
27.14.1 Decision
Change the claim to barrier coverage. The evidence record must review weak crossing segments, sensor orientation, active alarm state, maintenance ownership, and vegetation-change retest triggers. A high area-coverage percentage is not enough to accept this perimeter claim.
27.15 Implementation Checklist
Use this checklist before sending a coverage plan into deeper analysis.
Name the application goal Event, asset, region, boundary, route, or operating decision.
Name the coverage type Area, point, barrier, path, hybrid, or priority-zone mix.
Name the sensing model Boolean, probabilistic, measured, simulated, or hybrid.
Name the coverage level 1-coverage, k-coverage, priority-zone target, or degraded-state target.
Inspect operating state All-awake, active-set, rotation, mobile, maintenance, or degraded state.
Record acceptance limits Obstructions, calibration, gateway location, route state, battery state, season, owner, and retest trigger.
27.16 Common Mistakes
Starting with sensor count Counts are outputs of a claim and model. They should not define the coverage requirement.
Using one coverage type everywhere Area, point, barrier, and path coverage answer different questions.
Trusting nominal range Manufacturer or planning range is not field evidence. The record should state environmental limits.
Ignoring duty cycling The all-awake map can pass while the scheduled operating state fails.
Counting redundancy without independence Multiple sensors do not help if the same obstruction, relay failure, or power issue disables all of them.
Leaving ownership vague Coverage quality drifts unless someone owns retest, repair, and acceptance records.
27.17 Knowledge Check: Coverage Type
27.18 Knowledge Check: Operating State
27.19 Match the Coverage Planning Term
27.20 Order the Coverage Planning Workflow
27.21 Active-State Coverage Record
The body chapter names the coverage type, sensing model, redundancy target, active state, connectivity path, limits, owner, and retest trigger. The deeper rule is that those fields form one acceptance state, not a general promise about the whole network. A coverage record should answer which claim was accepted, which sensors were allowed to count, which evidence time was used, and which change would reopen the review.
Use the active-state map as the deeper check: a node only contributes live coverage when it senses the claim, remains in role, and can deliver evidence to the sink or local decision point.
A useful state record is closer to a small contract than a picture. It binds claim_type, geometry, sensing_model, coverage_level, active_set, delivery_path, evidence_time, and retest_trigger. If any one field changes, the old acceptance may still be historically true, but it is no longer proof of the current deployment. This is why all-awake planning maps, rotation schedules, degraded states, and mobile states need separate evidence records when their active nodes or delivery paths differ.
The same contract also prevents false comparisons between designs. A plan with a higher nominal coverage percentage can be weaker if it counts sleeping nodes, assumes an obsolete shelf layout, or omits the gateway path. A plan with a lower percentage can be stronger for the real mission if it names the priority zone, preserves the active set, and records exactly which uncovered region is outside the accepted claim. The state record is what lets those judgments be compared honestly.
Sensing state A sensor counts only if its model and evidence cover the claim in the reviewed geometry, not merely because it appears on a site drawing.
Role state Sleeping, disabled, uncalibrated, moved, or out-of-role nodes are named separately so they cannot silently inflate coverage.
Delivery state Evidence remains incomplete until an active node can deliver the reading through a live path or local decision rule.
27.22 Rehearse the Claim Against State Changes
A fundamentals review becomes practical when the team tests the accepted claim against the first failures that will happen in the field. The goal is not to imagine every possible accident. It is to separate decisions that stay true after a normal state change from decisions that must be narrowed, repaired, or retested. Each rehearsal should preserve the same fields as the record: claim, model, active set, delivery path, owner, and retest trigger.
| State change | Failure to test | Review response |
|---|---|---|
| Shelves or foliage move after acceptance. | The planned sensing model still assumes a clear path. | Narrow the claim until a walk test or calibration pass updates the evidence. |
| A rotation schedule puts two nodes to sleep. | The active set no longer provides the accepted redundancy level. | Review the weakest step and transition, not only the average schedule. |
| A relay or gateway is unreachable. | The region may be sensed but the evidence cannot leave the field. | Mark the state as sensed-but-undelivered unless local decision rules are accepted. |
| Two redundant sensors share a mount, battery, or obstruction. | The k-coverage label overstates independence. | Record the common-mode risk and lower the accepted redundancy or repair placement. |
The strongest field review language is usually explicit and limited. Write “accepted for rooms A, B, and C during the night rotation while gateway G1 is reachable” instead of “site covered.” Write “barrier segment three needs repair after vegetation growth” instead of “perimeter coverage failed.” That wording keeps the learner from treating coverage as a permanent badge. It also creates a maintainable queue: accept the parts with evidence, narrow the parts with uncertainty, repair the parts that fail, and retest the parts whose assumptions changed.
For coverage fundamentals, a good practitioner record is not longer than the deployment. It is just specific enough that the next reviewer can replay the acceptance state without guessing which nodes, model, and delivery path were trusted.
27.23 Sensing vs Delivery Predicates
The geometric shortcut most often taught in coverage fundamentals is useful only when its assumptions are visible. Let Rs be the sensing range and Rc the communication range. In a convex region with complete coverage by active nodes, Rc >= 2Rs is enough to make neighboring sensing disks communicate, so complete coverage implies network connectivity. The shortcut fails when the region is not convex, the coverage is partial, the active set changes, the radio range is obstructed, or the sink path depends on sleeping relays.
A robust implementation therefore stores two predicates. The sensing predicate asks whether location or target x is accepted by the chosen model at time t: C(x,t). The delivery predicate asks whether the active node that supports that claim can deliver evidence to a sink, gateway, or accepted local decision point: D(s,t). A coverage claim is operationally true only where the accepted sensing predicate and delivery predicate are both true for the same state. This avoids a common error: reporting a beautiful coverage ratio while the supporting readings are stranded behind a sleeping relay.
In practice, those predicates are evaluated over a finite review grid, named targets, boundary samples, or path segments. Each sample should carry the sensor or sensors that satisfy C, the route or local rule that satisfies D, and the interval for which both remain true. That interval matters: a node can satisfy sensing during a sample window and lose delivery during a later store-forward gap. Treating the two predicates as timestamped graph checks keeps the coverage record honest when duty cycling, gateway movement, and maintenance windows change the graph.
When the shortcut holds Record the convex-region, complete-coverage, active-node, and range assumptions, then treat coverage as the harder constraint.
When it does not Verify sensing coverage and delivery connectivity separately, especially during rotation, degradation, and gateway movement.
27.23.1 Knowledge Check: Sensing and Delivery
27.24 Summary
WSN coverage fundamentals are about disciplined claim framing. Before choosing algorithms or counting sensors, state the monitoring goal, coverage type, sensing model, redundancy level, operating state, connectivity path, limits, owner, and retest trigger. This keeps the coverage plan meaningful as the deployment moves from drawing to field operation.
27.25 Key Takeaway
WSN Coverage: Fundamentals should connect coverage goals to sensing model, placement density, redundancy, obstacles, energy use, mobility, and deployment evidence.
27.26 Concept Relationships
- WSN Coverage Problem Types expands area, point, barrier, and path coverage choices.
- WSN Coverage Algorithms helps select verification, active-set, rotation, and gap-repair algorithm families.
- WSN Coverage Worked Examples applies the planning route to concrete field scenarios.
- WSN Coverage Implementations turns accepted coverage claims into placement, active-set, validation, and retest decisions.
- WSN Energy Duty Cycling explains how sleep schedules affect active coverage states.
27.27 What’s Next
- WSN Coverage Problem Types for selecting the right coverage formulation.
- WSN Coverage Algorithms for verification, active-set, rotation, and gap-repair algorithm families.
- WSN Coverage Worked Examples for applying the planning workflow to concrete scenarios.
- WSN Coverage Implementations for implementation patterns after the fundamentals are clear.