244  Ad-Hoc Networks: Applications and Practice

244.1 Learning Objectives

By the end of this chapter, you will be able to:

  • Identify Use Cases: Recognize when ad-hoc networks are the appropriate solution for IoT deployments
  • Apply Routing Selection: Choose appropriate routing protocols for specific deployment scenarios
  • Calculate Energy Trade-offs: Analyze multi-hop vs direct transmission energy costs
  • Design for Scale: Understand scalability limitations and solutions
  • Test Your Knowledge: Validate understanding through worked examples and quizzes

244.2 Prerequisites

Before diving into this chapter, you should be familiar with:

244.3 Ad-Hoc Network Applications

⏱️ ~10 min | ⭐⭐ Intermediate | 📋 P04.C02.U04

⭐⭐ Intermediate

%% fig-alt: "Real-world ad-hoc network applications showing five deployment scenarios: disaster recovery with rescue teams forming mesh network over destroyed infrastructure, military tactical operations with encrypted mobile communications, vehicular networks with V2V safety message exchange, wireless sensor networks with environmental monitoring data relay, and flying ad-hoc networks with drone swarms coordinating via aerial mesh"
%%{init: {'theme': 'base', 'themeVariables': {'primaryColor': '#2C3E50', 'primaryTextColor': '#fff', 'primaryBorderColor': '#16A085', 'lineColor': '#E67E22', 'secondaryColor': '#7F8C8D', 'tertiaryColor': '#ECF0F1'}}}%%
graph TB
    subgraph Disaster["Disaster Recovery"]
        DR1[Destroyed<br/>Infrastructure]
        DR2[Rescue Team<br/>Radios]
        DR3[Ad-hoc<br/>Mesh]
        DR1 --> DR2 --> DR3
    end

    subgraph Military["Military Tactical"]
        MT1[Mobile<br/>Soldiers]
        MT2[Encrypted<br/>Comms]
        MT3[Dynamic<br/>Topology]
        MT1 --> MT2 --> MT3
    end

    subgraph VANET["Vehicular Networks"]
        VN1[Moving<br/>Vehicles]
        VN2[V2V<br/>Messages]
        VN3[Safety<br/>Alerts]
        VN1 --> VN2 --> VN3
    end

    subgraph WSN["Sensor Networks"]
        WS1[Remote<br/>Sensors]
        WS2[Multi-hop<br/>Relay]
        WS3[Gateway<br/>Upload]
        WS1 --> WS2 --> WS3
    end

    subgraph FANET["Drone Swarms"]
        FA1[UAV<br/>Formation]
        FA2[Aerial<br/>Mesh]
        FA3[Coordinated<br/>Mission]
        FA1 --> FA2 --> FA3
    end

    style DR3 fill:#E67E22,stroke:#2C3E50,color:#fff
    style MT3 fill:#E67E22,stroke:#2C3E50,color:#fff
    style VN3 fill:#E67E22,stroke:#2C3E50,color:#fff
    style WS3 fill:#E67E22,stroke:#2C3E50,color:#fff
    style FA3 fill:#E67E22,stroke:#2C3E50,color:#fff

Figure 244.1: Real-world ad-hoc network applications showing five deployment scenarios: disaster recovery, military tactical, vehicular networks, sensor networks, and drone swarms.

244.3.1 Disaster Recovery and Emergency Response

⭐⭐ Intermediate

  • Scenario: Earthquake destroys cellular towers; rescue teams need communication
  • Solution: Handheld radios form ad-hoc mesh; relay messages multi-hop
  • Requirements: Rapid deployment, high reliability, no infrastructure dependency
  • Real Deployment: 2011 Fukushima disaster - rescue teams deployed 150 ad-hoc radios covering 25km² with 4-6 hop paths, enabling coordination when all cellular networks failed for 72 hours

244.3.2 Military and Tactical Operations

  • Scenario: Soldiers in battlefield need real-time coordination
  • Solution: Wearable radios create mobile mesh network; encrypted communications
  • Requirements: High security, low latency, resilience to node capture

244.3.3 Vehicular Ad-Hoc Networks (VANETs)

  • Scenario: Cars exchange traffic, safety, and infotainment data
  • Solution: 802.11p/DSRC radios create vehicle-to-vehicle (V2V) mesh
  • Requirements: High mobility support, low latency (collision warnings), privacy

244.3.4 Wireless Sensor Networks (WSNs)

  • Scenario: Forest fire detection using temperature/smoke sensors
  • Solution: Battery-powered sensors relay data multi-hop to gateway
  • Requirements: Energy efficiency, scalability (1000s of nodes), long lifetime

244.3.5 Flying Ad-Hoc Networks (FANETs)

  • Scenario: Drone swarm for search-and-rescue or surveillance
  • Solution: Drones relay video and coordinates via ad-hoc mesh
  • Requirements: 3D topology support, high mobility, GPS integration

244.4 Worked Examples

NoteWorked Example: Selecting Routing Protocol for Wildlife Tracking Network

Scenario: A conservation team is deploying GPS collars on 50 elephants across a 200 km² wildlife reserve in Kenya. Collars transmit location data twice daily to solar-powered relay nodes scattered throughout the reserve. The relay nodes form an ad-hoc mesh to route data to a central base station.

Given:

  • 50 elephant collars (mobile, unpredictable movement)
  • 30 relay nodes (stationary, solar-powered)
  • Communication frequency: 2 transmissions per collar per day = 100 total transmissions/day
  • Network diameter: 8-10 hops from furthest relay to base station
  • Latency tolerance: 5 minutes (non-real-time)

Steps:

  1. Analyze traffic pattern: 100 transmissions/day = ~4 transmissions/hour = 1 every 15 minutes. This is sparse traffic - nodes are mostly idle.

  2. Evaluate mobility: Relay nodes are stationary (good for route stability). Elephant collars move but only communicate twice daily, so routes can be discovered fresh each time.

  3. Calculate proactive overhead: DSDV with 30 relays, 15-second updates = 30 nodes x 4 updates/minute = 120 control packets/minute = 7,200/hour. With only 4 data packets/hour, overhead ratio is 1800:1 (unacceptable).

  4. Calculate reactive overhead: DSR with 100 discoveries/day, ~50 packets per discovery = 5,000 control packets/day vs. 100 data packets/day. Overhead ratio is 50:1 (acceptable for solar-powered nodes).

  5. Consider latency: 5-minute tolerance easily accommodates DSR’s 500ms-2s discovery delay.

Result: Reactive routing (DSR) is optimal. The sparse communication pattern (100 transmissions/day) makes proactive overhead prohibitive. DSR’s discovery delay (1-2 seconds) is negligible compared to the 5-minute latency tolerance.

Key Insight: For IoT deployments with infrequent communication (<1 transmission per node per hour), reactive routing typically provides 10-100x lower overhead than proactive approaches. The key decision factor is the ratio of communication frequency to routing update frequency.

NoteWorked Example: Calculating Multi-Hop Energy Savings

Scenario: A smart agriculture deployment monitors soil moisture across a 500m x 500m field. A sensor at the corner (500m from the gateway) can either transmit directly using high-power mode or use multi-hop through intermediate sensors spaced 100m apart.

Given:

  • Direct transmission distance: 500m (diagonal)
  • Multi-hop path: 5 hops of 100m each
  • Direct transmission power: 100 mW (to achieve 500m range)
  • Multi-hop transmission power: 2 mW per hop (for 100m range)
  • Packet transmission time: 10 ms
  • Relay overhead: 5 ms processing per hop for receive + forward
  • Sensor battery: 2400 mAh (2x AA batteries)
  • Transmission voltage: 3.3V

Steps:

  1. Calculate direct transmission energy:
    • Power: 100 mW at 3.3V = 30.3 mA
    • Duration: 10 ms
    • Energy: 30.3 mA x 10 ms = 0.303 mAs = 84.2 nWh per packet
  2. Calculate multi-hop source energy (just the originating sensor):
    • Power: 2 mW at 3.3V = 0.61 mA
    • Duration: 10 ms
    • Energy: 0.61 mA x 10 ms = 6.1 uAs = 1.69 nWh per packet
  3. Calculate total network energy (all 5 hops):
    • Each hop: transmit (10 ms x 2 mW) + receive (10 ms x 15 mW typical) + process (5 ms x 30 mW)
    • Per hop: 0.02 + 0.15 + 0.15 = 0.32 mJ
    • Total 5 hops: 5 x 0.32 mJ = 1.6 mJ vs. direct: 1.0 mJ
  4. Evaluate trade-off:
    • Source sensor savings: 84.2 nWh -> 1.69 nWh = 98% reduction
    • Total network energy: Multi-hop uses 1.6x more total energy
    • But energy is distributed across 5 nodes instead of concentrated at one

Result: Multi-hop routing reduces the corner sensor’s energy consumption by 98% (from 84.2 nWh to 1.69 nWh per packet). While total network energy increases 60%, the energy is distributed across multiple nodes, preventing premature battery death at the corner sensor.

Key Insight: Multi-hop routing doesn’t minimize total energy—it distributes energy across multiple nodes. This prevents “energy holes” where sensors near the gateway or at network edges deplete first. For battery-powered networks, energy distribution is often more important than total efficiency.

244.5 Knowledge Check

Test your understanding of these architectural concepts.

Question 1: A battery-powered sensor network monitors soil moisture across a 100-hectare farm. Sensors transmit readings once per hour to a central gateway. Each sensor has 2 neighbors on average, and the network diameter is 8 hops. Which routing protocol class is most appropriate?

Explanation: Reactive routing (DSR) is optimal here due to sparse communication pattern. Analysis: Sensors transmit once per hour = 24 transmissions/day. DSDV overhead: 100 nodes x updates every 15s = 576,000 control packets/day. DSR overhead: 24 route discoveries x ~50 packets/discovery = 1,200 packets/day. DSR uses 480x less control overhead! Battery savings: DSDV requires nodes to stay awake for periodic updates (drains battery). DSR allows deep sleep between hourly transmissions. Why not ZRP? Network is sparse (2 neighbors avg) - zones would contain few nodes, eliminating hybrid benefit. Why not static? Node failures require manual reconfiguration. DSR auto-discovers alternate paths. Trade-off accepted: DSR adds ~500ms discovery delay before first transmission, acceptable for soil moisture (non-time-critical).

Question 2: In DSDV, why are sequence numbers critical for route selection? What problem do they solve that hop count alone cannot?

Explanation: Sequence numbers solve the “stale route” problem that causes routing loops. Scenario without sequence numbers: Node A has route to D via B (2 hops). Link B-D breaks. Node C still advertises old route: “D via B, 3 hops”. Node A receives this and installs it (lower hop count than infinity). Result: A sends packets to B which loops! With sequence numbers: When B-D breaks, D increments its sequence number (seq=51). B advertises broken route with new seq=51 and hop=infinity. Old route from C (seq=50) is rejected because seq 50 < seq 51. Rule: Higher sequence number = fresher information. DSDV always prefers fresher routes, even if longer. Only when sequences match does hop count decide. This ensures broken route information propagates faster than outdated “good” routes.

Question 3: A sensor 500m from the gateway can transmit directly at 100mW or use 5-hop multi-hop at 5mW per hop. Which is more energy-efficient for the transmitting sensor?

Explanation: For the source sensor, multi-hop is 20x more efficient (5mW vs 100mW). However, this question highlights a common confusion. Source energy: Direct = 100mW x 10ms = 1mJ. Multi-hop = 5mW x 10ms = 0.05mJ. Source saves 95% energy! Network total: Multi-hop requires 4 relay nodes to also transmit. Each relay: 5mW x 10ms = 0.05mJ. Total network: 5 x 0.05mJ = 0.25mJ. Comparison: Direct = 1mJ (source only). Multi-hop = 0.25mJ (distributed across 5 nodes). Multi-hop is still 4x more efficient overall because radio power scales with distance squared (Friis equation). Key insight: Multi-hop distributes energy cost across many nodes, reducing individual node burden and preventing “energy holes” near the gateway.

Question 4: In Zone Routing Protocol (ZRP), what happens if the zone radius is set larger than the network diameter?

Explanation: When zone radius is greater than or equal to network diameter, all nodes are within every node’s zone. Result: IARP (intra-zone proactive) covers entire network. IERP (inter-zone reactive) never triggered - no destinations “outside” zone. ZRP = pure DSDV with full proactive overhead! Example: 50-node network, diameter 8 hops, zone radius=10. Every node maintains routes to all 50 nodes continuously. Overhead: 50 x periodic updates = same as DSDV. Hybrid benefit lost: No reactive savings for distant destinations. Optimal zone radius: Typically 2-3 hops, covering ~15-30% of network. This balances: Local traffic uses pre-computed routes (low latency), Distant traffic uses on-demand discovery (low overhead). Rule of thumb: Zone radius should be much smaller than network diameter to leverage hybrid benefits.

244.7 Summary

This chapter explored the practical applications of ad-hoc networks and provided worked examples to solidify understanding of routing protocol selection and energy trade-offs.

244.7.1 Key Takeaways

  1. Diverse Applications: Ad-hoc networks are essential for disaster recovery, military operations, vehicular networks (VANETs), wireless sensor networks (WSNs), and drone swarms (FANETs).

  2. Protocol Selection Depends on Traffic:

    • Sparse, infrequent communication -> Reactive routing (DSR, AODV)
    • Dense, continuous communication -> Proactive routing (DSDV, OLSR)
    • Large, heterogeneous networks -> Hybrid routing (ZRP)

  3. Energy Distribution Matters: Multi-hop routing doesn’t minimize total energy—it distributes energy across nodes, preventing premature battery death at edge nodes.

  4. Overhead Calculations: For sparse traffic (<1 transmission/node/hour), reactive routing provides 10-100x lower overhead than proactive approaches.

  5. Scalability Planning: Design for 3-5x initial node count; flat routing struggles beyond 100-200 nodes.

244.7.2 Design Guidelines Summary

  • Use proactive routing when: Network is small (<50 nodes), relatively static, and has continuous traffic patterns
  • Use reactive routing when: Network is large, mobile, energy-constrained, and has bursty or sparse traffic
  • Use hybrid routing when: Network exceeds 100 nodes with heterogeneous mobility and traffic patterns
  • Prefer multi-hop when: Coverage area exceeds single-hop radio range (typically >100m for low-power IoT radios)
  • Consider infrastructure when: Deployment is permanent, bandwidth requirements are high (>1 Mbps), or ultra-low latency (<50ms) is critical

244.8 What’s Next

You’ve now completed the ad-hoc network fundamentals series. Continue with deep dives into specific protocols:

244.8.1 Deep Dive into Routing Protocols

244.8.2 Advanced Topics

244.8.4 Protocols Building on Ad-Hoc Principles

244.8.5 Hands-On Learning

Recommended Next: Ad-hoc Routing: Proactive (DSDV) to understand table-driven routing in detail.