“I need to send a message to a sensor on the other side of the farm,” said Sammy. “But the farm is 1 kilometer wide and my radio only reaches 100 meters! I need to hop through other sensors to get there.”
Max the Microcontroller explained three strategies:
“Strategy 1 - Proactive (Know Everything): Keep a map of where EVERY sensor is, updated every few seconds. When you need to send a message, you already know the way! Fast but exhausting – it’s like constantly updating a map of every friend’s location even when you don’t need to talk to them.”
“Strategy 2 - Reactive (Ask When Needed): Don’t keep any map. When you need to send a message, shout ‘WHO KNOWS HOW TO REACH SENSOR #42?’ and wait for an answer. Saves energy when you’re quiet, but takes time to find the path!”
“Strategy 3 - Hybrid (Best of Both): Keep a map of your NEIGHBORHOOD (nearby sensors) and ask for directions only when reaching far-away sensors. Like knowing your own street by heart but using a GPS for cross-town trips!”
Bella the Battery added, “I like Strategy 3 best – it doesn’t drain me keeping track of the whole farm, but I still know my local friends!”
Why multi-hop? Low-power IoT radios (Zigbee, BLE) reach only 10-100 meters, but deployment areas span kilometers. Messages must “hop” through intermediate devices.
The three approaches:
| Approach | How It Works | Best For |
|---|---|---|
| Proactive | Keep routes to everywhere, always | Small stable networks with frequent traffic |
| Reactive | Find routes only when needed | Large mobile networks with rare traffic |
| Hybrid | Local routes always ready; distant routes on-demand | Large networks with mixed traffic patterns |
Energy insight: Sending directly over 500m uses ~100mW. Sending in 5 hops of 100m uses ~2mW per hop. The source sensor saves 98% energy with multi-hop because radio power scales with the square of distance!
Understanding how the three routing paradigms actually operate reveals their fundamental trade-offs. Here’s a step-by-step walkthrough comparing proactive, reactive, and hybrid routing for the same communication scenario.
Scenario Setup:
Network of 12 nodes arranged in a 3×4 grid, each node can directly communicate with its 4 neighbors (N, S, E, W). Node A (top-left corner) wants to send a data packet to Node L (bottom-right corner), requiring a 5-hop path (A→B→C→G→K→L).
Pre-Communication Phase (Continuous Maintenance):
Step 1: Initial Table Construction
Step 2: Routing Table Exchange (Periodic Updates)
Step 3: Sequence Number Management
Data Transmission Phase:
Step 4: Zero-Latency Forwarding
Overhead Summary (24-Hour Period):
Pre-Communication Phase:
Step 1: Minimal Initialization
Data Transmission Phase:
Step 2: Route Discovery (Flooding)
Step 3: Route Reply
Step 4: Data Transmission Using Cached Route
Step 5: Route Cache Timeout
Overhead Summary (24-Hour Period):
Pre-Communication Phase:
Step 1: Zone Construction
Step 2: Intra-Zone Proactive Routing
Data Transmission Phase:
Step 3: Determine Intra-Zone vs Inter-Zone
Step 4: Bordercast Route Discovery
Step 5: Hybrid Path Construction
Overhead Summary (24-Hour Period):
Comparison Timeline for One Data Packet (A → L):
| Time | Proactive (DSDV) | Reactive (DSR) | Hybrid (ZRP) |
|---|---|---|---|
| t=0 | Packet arrives at A | Packet arrives at A | Packet arrives at A |
| t=0 | Route ready (from periodic updates) | No route - start discovery | Partial route (intra-zone only) |
| t=5ms | Forwarded to B | RREQ flooded | Intra-zone forwarding to C |
| t=100ms | - | RREQ reaches L | Bordercast RREQ sent via C |
| t=200ms | - | RREP reaches A | RREP reaches A |
| t=225ms | Packet delivered (25ms total) | Packet delivered (225ms total) | Packet delivered (125ms total) |
Key Insights:
Protocol Selection Rule from This Walkthrough:
By the end of this chapter, you will be able to:
Before diving into this chapter, you should be familiar with:
⭐⭐ Intermediate
Wireless range limitations necessitate multi-hop communication:
Multi-hop energy savings via Friis equation: Radio power scales with distance squared. For direct transmission over 500m requiring 100mW power, splitting into 5 hops of 100m each requires only 4mW per hop (since (100m)² = 10,000 vs. (500m)² = 250,000 → 25× difference). Worked example: Source sensor saves 96% energy (100mW → 4mW transmit), but total network energy increases (5 nodes × 4mW = 20mW), distributing the load and preventing premature battery death at any single sensor.
Key routing challenges in ad-hoc networks:
⭐⭐⭐ Advanced
Concept: Nodes maintain routes to all destinations at all times, even before packets need sending.
How It Works:
Examples: DSDV (Destination-Sequenced Distance Vector), OLSR (Optimized Link State Routing)
Trade-offs:
Best For: Dense, relatively static networks with continuous traffic
Concept: Routes discovered only when needed; no periodic updates.
How It Works:
Examples: DSR (Dynamic Source Routing), AODV (Ad-hoc On-Demand Distance Vector)
Trade-offs:
Best For: Sparse, mobile networks with bursty traffic
Concept: Combines proactive (within zones) and reactive (between zones) approaches.
How It Works:
Examples: ZRP (Zone Routing Protocol), ZHLS (Zone-based Hierarchical Link State)
Trade-offs:
Best For: Large, heterogeneous networks with varying traffic patterns
This variant shows the same routing protocol selection as a layered architecture - emphasizing how application requirements (top) and network characteristics (bottom) influence routing strategy selection in the middle layer, which then maps to specific protocol implementations:
This variant shows the same routing approaches as a side-by-side comparison - emphasizing the operational differences between proactive and reactive routing strategies and their trade-offs:
| Metric | Proactive | Reactive | Hybrid |
|---|---|---|---|
| Route Discovery Latency | Low (immediate) | High (flooding delay) | Medium (zone-dependent) |
| Routing Overhead | High (periodic updates) | Low (on-demand) | Medium (zone updates) |
| Scalability | Poor (O(n²) updates) | Good (O(n) per route) | Good (O(zone size)) |
| Memory Usage | High (full routing table) | Low (active routes only) | Medium (zone + cache) |
| Energy Efficiency | Low (constant overhead) | High (minimal updates) | Medium (zone-dependent) |
| Best Use Case | Dense, static, continuous traffic | Sparse, mobile, bursty traffic | Large, heterogeneous |
Scenario: A forest fire detection network has sensors in a line, each 100m apart, extending 1km from gateway. Direct transmission requires 100 mW power but only reaches 100m. Multi-hop transmission uses 10 mW per hop (10x more energy-efficient per hop, but requires multiple hops).
Think about:
Key Insight: Multi-hop is more efficient for distances > 2-3 hops.
Energy comparison:
Break-even point: Multi-hop becomes efficient when: Hops x E_hop < E_direct. For this example: Hops x 0.1 mJ < 1 mJ -> Hops < 10.
Correction: Radio power scales with distance squared (Friis equation). For 1 km direct transmission, power might be 1000 mW (not 100 mW). Then:
Multi-hop is 10x more efficient!
Complicating factors:
Design guideline: Multi-hop is energy-efficient for battery-powered networks spanning >500m, but adds latency and complexity. Hybrid approaches use high-power direct links for critical nodes and multi-hop for non-critical sensors.
The mistake: Deploying reactive routing protocols (DSR, AODV) in networks with regular, continuous data flows - such as sensors reporting every 30 seconds.
Why it happens: Teams choose reactive routing because it “saves energy when idle,” without analyzing actual traffic patterns. Marketing materials emphasize on-demand efficiency without mentioning the break-even point.
The fix: Calculate your traffic pattern first. If nodes send data more frequently than once every 2-5 minutes, proactive routing (DSDV, OLSR) typically consumes less energy because route discovery flooding overhead exceeds periodic update cost. Use reactive routing only for sparse, event-driven traffic (alarms, motion detection).
The mistake: Deploying ad-hoc networks without accounting for hidden terminals - nodes that cannot hear each other but interfere at a common receiver, causing collisions and retransmissions.
Why it happens: Simulation tools often use simplified propagation models where “in range” means “can communicate.” Real deployments have asymmetric links, obstacles, and interference patterns that create hidden node scenarios invisible during testing.
The fix: Enable RTS/CTS (Request-to-Send/Clear-to-Send) for networks with >10 nodes or irregular topologies. Use site surveys to identify hidden node pairs. Consider carrier sense threshold tuning. Budget for 20-30% more retransmissions than simulations predict.
The mistake: Using flat routing protocols (where all nodes are equal peers) in networks exceeding 100-150 nodes, leading to routing table explosion and control message flooding.
Why it happens: Initial deployments work fine at 20-50 nodes. As the network grows organically, teams don’t re-evaluate routing architecture. The scalability wall hits suddenly - network performance degrades exponentially, not linearly.
The fix: Design for 3-5x your initial node count from day one. For networks expecting >100 nodes, use hierarchical routing (cluster-based like LEACH) or hybrid approaches (ZRP). Monitor routing table sizes and control message ratios - if control traffic exceeds 15% of total bandwidth, it’s time to restructure.
Question 1: A forest fire detection network has sensors reporting temperature every 30 seconds to a central gateway. Which routing approach is most appropriate?
b) Proactive routing. With sensors reporting every 30 seconds, traffic is frequent and continuous. Reactive routing would waste energy rediscovering routes every 30 seconds (route discovery flooding overhead exceeds proactive periodic updates). Proactive routing amortizes its update cost across the many data transmissions. The rule of thumb: if transmission rate > 1 packet every 2-5 minutes, proactive wins.
Question 2: Why does transmit power scale with the square of distance in wireless communications (Friis equation), and what implication does this have for multi-hop routing?
b) Multiple short hops use dramatically less source energy than one long hop. The Friis equation shows that transmit power scales with distance squared. To reach 500m directly requires ~100mW, but five 100m hops require only ~2mW each. The source sensor saves 98% energy. While total network energy may increase (5 relays each use some energy), the energy is distributed across nodes, preventing premature battery death at any single sensor.
Question 3: What is the primary scalability limitation of flat ad-hoc routing protocols?
b) Routing table size and control overhead grow quadratically. In proactive protocols, each of N nodes maintains routes to N-1 destinations, requiring O(N^2) total routing entries network-wide. Control messages grow similarly. Flat ad-hoc routing typically struggles beyond 100-200 nodes. Solutions include hierarchical routing (cluster heads), geographic routing (location-based), or hybrid approaches like ZRP.
Scenario: Compare control message overhead for DSDV (proactive) vs DSR (reactive) in 100-node industrial monitoring network over 24-hour period.
Given:
Steps:
DSDV Overhead:
DSR Overhead:
Comparison:
| Protocol | Overhead (MB/day) | vs Data Traffic | Winner |
|---|---|---|---|
| DSDV | 1,360.57 | 25× more than DSR | ❌ High overhead |
| DSR | 54.47 | 96% less than DSDV | ✓ Low overhead for sparse traffic |
| Data | 28,800 packets × 100 bytes = 2.88 MB/day | - | Baseline |
DSDV overhead is 472× the actual data traffic! Clearly unsuitable for this sparse-traffic scenario.
Result: DSR wins dramatically for this deployment (54 MB vs 1,361 MB control overhead). DSDV’s continuous updates waste 96% more bandwidth despite serving the same traffic.
Key Insight: The break-even point depends on traffic frequency. If nodes sent 1 packet every 10 seconds (vs 5 minutes), discoveries would occur constantly, and DSR overhead would exceed DSDV. Rule of thumb: Use DSR when transmission interval > 2-3 minutes; use DSDV when interval < 30 seconds.
Use measurable network parameters to select the optimal routing paradigm.
| Parameter | Proactive (DSDV) | Reactive (DSR) | Hybrid (ZRP) |
|---|---|---|---|
| Packet rate per node | >2 pkt/min (frequent) | <0.5 pkt/min (sparse) | 0.5-2 pkt/min (mixed) |
| Network size | <50 nodes | 50-500 nodes | 100-1000 nodes |
| Topology stability | Link change <1/hour | Link change 1-20/hour | Link change 0.5-10/hour |
| Average path length | <4 hops | 4-10 hops | 5-15 hops |
| Latency requirement | <100ms critical | >500ms tolerable | <300ms for 80% traffic |
| Energy budget | Mains/abundant solar | Battery, 2-5 year life | Battery with periodic recharge |
| Memory per node | 10-50 KB routing tables | 2-10 KB route cache | 5-20 KB (zone + cache) |
| Break-even calc | Update_cost < Discovery_cost × Pkt_rate | Discovery_cost < Update_cost / Pkt_rate | Tune zone radius |
Break-Even Formula:
When is Proactive Better?
Proactive overhead/day < Reactive overhead/day
(N × update_size × updates_per_day) < (discoveries_per_day × flood_cost)
Where:
- discoveries_per_day = packets_per_day × (1 - cache_hit_rate) × unique_destinations / N
- flood_cost = N × RREQ_size + path_length × RREP_sizeExample Calculation:
But if packets increase to 1/second:
Hybrid ZRP Tuning:
Key Insight: No universal winner - protocol selection depends on your specific traffic pattern, not protocol “goodness.”
The Mistake: Deploying DSR or AODV for IoT sensor networks that report every 30-60 seconds, causing continuous route discovery storms that waste more energy than data transmission.
Real-World Example: A 2019 smart factory deployment used AODV for 80 vibration sensors reporting every 45 seconds. After 2 weeks, operators noticed: - Average discovery delay: 180ms per packet - Route discoveries: 15,360/hour network-wide - Control traffic: 87% of all transmissions - Battery drain: 3× higher than expected
Why This Happens:
Misconception: “Reactive routing saves energy by not doing periodic updates”
Reality: Reactive routing saves energy only when traffic is sparse (several minutes between packets). For frequent traffic, route discovery overhead exceeds proactive maintenance.
Quantified Analysis:
Given: 80 sensors, 45-second reporting interval, 5-hop average path, route cache 5-minute timeout
Wait - DSDV overhead (37.92 MB) >> DSR overhead (4.35 MB)? So DSR still wins?
Missing Factor: Forwarding Overhead
Verdict: DSR still wins for this scenario… BUT:
Real Problem: Route Error Storms in Industrial Environment
Still less than DSDV’s 37.92 MB! So why did deployment fail?
The Real Killer: Discovery Latency + Retransmissions
Solution: Switch to DSDV:
Key Lesson: Don’t choose reactive routing just because “reactive = lower overhead.” For frequent traffic (>1 pkt/2min) or latency-critical applications, proactive routing’s overhead is justified by zero discovery delay and no RERR storms.
Design Rule: Use reactive routing when: - Traffic interval > 3× route cache timeout - Latency tolerance > 500ms - Environment is stable (low link break rate)
Otherwise, use proactive or hybrid.
Apply the routing protocol selection criteria from this chapter to a realistic vehicle fleet scenario. Work through the analysis step-by-step, then compare your solution with the provided analysis.
Scenario:
A construction company operates 50 trucks across a 20 km² construction site. Each truck is equipped with: - GPS receiver - Short-range radio (802.11p DSRC, 300m range, 6 Mbps) - IoT gateway collecting sensor data (engine diagnostics, load weight, fuel level)
Communication Requirements:
Your Design Task:
Choose a routing protocol from these options:
Option A: Proactive (DSDV)
Option B: Reactive (DSR)
Option C: Hybrid (ZRP)
Option D: Geographic (GPSR)
Questions to Answer:
Work Through Your Solution:
Before checking the analysis below, write down: - Your chosen routing protocol (A, B, C, or D) - Key factors that influenced your decision - Expected trade-offs (what you’re gaining vs. sacrificing)
Recommended Solution: Option D (Geographic Routing - GPSR)
Justification Through Systematic Analysis:
1. Traffic Pattern Breakdown:
Broadcast Traffic (Status Updates):
Unicast Traffic (Coordinator ↔︎ Trucks):
Key Insight #1: Status updates dominate bandwidth (43.2 MB) but don’t need routing. Only 48 KB/day requires multi-hop routing. This is sparse unicast traffic despite high total packet volume.
2. Overhead Calculation for Each Protocol:
Option A - Proactive (DSDV):
Option B - Reactive (DSR):
Option C - Hybrid (ZRP, 2-hop zones):
Option D - Geographic (GPSR):
3. Topology Dynamics (Mobility Challenge):
Impact on Topology-Based Routing:
Impact on Geographic Routing:
4. Latency Requirements:
Latency for Each Protocol:
All protocols meet the 500ms deadline, but GPSR and DSDV have lowest latency.
5. Decision Matrix:
| Criterion | DSDV | DSR | ZRP | GPSR | Weight |
|---|---|---|---|---|---|
| Overhead | ✗ 1.08 GB/day | ✓ 1.04 MB/day | ⚠ 216 MB/day | ✓✓ 0 MB | High |
| Latency | ✓✓ 40ms | ⚠ 340ms | ✓ 190ms | ✓✓ 40ms | High |
| Mobility resilience | ✗ 7.5 MB/hour updates | ✗ 6.5 MB/hour rediscoveries | ⚠ 5 MB/hour | ✓✓ Immune (GPS-based) | Critical |
| Scalability | ✗ O(n²) table size | ✓ O(1) storage | ⚠ O(zone size) | ✓✓ O(1) storage | Medium |
| Implementation | ✓ Standard protocol | ✓ Standard protocol | ⚠ Complex zones | ✓ Requires GPS (already present) | Low |
Winner: GPSR (Geographic Routing)
Reasons:
Trade-offs Accepted:
Alternative Valid Answer: DSR (Reactive Routing)
If GPS is unavailable or unreliable, DSR is the fallback choice: - 1.04 MB/day overhead is 1,000× less than DSDV - 340ms latency still meets the 500ms emergency alert deadline - 6.5 MB/hour rediscovery overhead (from mobility) is manageable at 6 Mbps data rate (0.3% bandwidth utilization)
Why NOT DSDV or ZRP:
DSDV: 1.08 GB/day overhead consumes 12% of available bandwidth (6 Mbps × 86,400 sec = 64.8 GB/day capacity). For 48 KB/day of unicast traffic, this is absurdly wasteful. Mobility causes constant routing storms (7.5 MB/hour triggered updates).
ZRP: 216 MB/day overhead is 200× better than DSDV but 200× worse than reactive. Hybrid is designed for mixed traffic (some dense, some sparse), but this scenario has uniformly sparse unicast traffic (48 KB/day). All the intra-zone overhead is wasted.
What to Observe After Deployment:
Monitor these metrics to validate the GPSR choice:
Greedy forwarding success rate: Should be >95% (perimeter mode <5% of the time). If perimeter mode exceeds 10%, topology is too sparse or obstacles are too numerous - add more trucks or switch to DSR.
Delivery latency: 95th percentile should be <100ms for coordinator-to-truck messages. If >200ms, check GPS update frequency (may need to increase from 5sec to 2sec).
Bandwidth utilization: Status broadcast (43.2 MB/day) + routing overhead (ideally <1 MB/day) should consume <1% of 6 Mbps capacity. If >5%, check for broadcast storms or GPS beacon duplication.
Emergency alert delivery rate: Should be 100% within 500ms. If <98%, investigate dead zones (GPS blackout areas) or perimeter routing failures.
Connection to Chapter Concepts:
This exercise applied all three routing paradigms and revealed when geography-based routing outperforms topology-based approaches:
Key Lesson: Don’t default to “reactive = best for sparse traffic.” If nodes already have location information (GPS, indoor positioning), geographic routing eliminates routing overhead entirely while maintaining low latency. This is why vehicular ad-hoc networks (VANETs) almost universally use geographic routing (GPSR, GeoDTN).
Multi-hop routing is the backbone of ad-hoc networks, enabling communication across distances far beyond single-hop radio range. This chapter covered the three fundamental routing paradigms and their trade-offs.
Multi-Hop Necessity: Wireless range limitations (10-100m for low-power radios) require multi-hop routing to cover deployment areas spanning hundreds of meters to kilometers.
Three Routing Paradigms:
Critical Trade-Offs:
Protocol Selection Guidelines:
Energy Considerations: Multi-hop routing distributes energy consumption across nodes, preventing “energy holes” near the gateway.
Estimate daily control overhead for proactive vs reactive routing based on your network parameters:
Understanding how multi-hop routing concepts interconnect helps build intuition for protocol selection and network design. The table below maps key relationships between concepts covered in this chapter.
| Concept A | Relationship | Concept B | Explanation |
|---|---|---|---|
| Multi-Hop Routing | necessary due to | Limited Radio Range | Low-power IoT radios reach only 10-100m, but deployment areas span kilometers. Multi-hop extends coverage by relaying through intermediate nodes. |
| Multi-Hop Routing | trades | Energy Savings vs. Latency | Transmit power scales with distance² (Friis equation), so 5 short hops use 98% less source energy than one long hop. But each hop adds 5-20ms latency and relay forwarding overhead. |
| Proactive Routing | maintains | Complete Routing Tables | Each node stores routes to ALL destinations via periodic table exchanges (e.g., DSDV broadcasts every 15 sec), enabling immediate forwarding when data arrives. |
| Proactive Routing | provides | Zero Discovery Latency | Routes already exist before data transmission begins. Packets forwarded immediately (0ms discovery) at the cost of continuous update overhead. |
| Proactive Routing | suited for | Dense/Continuous Traffic | When nodes communicate frequently (>1 packet every 2-5 minutes), the amortized cost of periodic updates is less than repeated on-demand discoveries. |
| Reactive Routing | discovers routes | On-Demand via Flooding | When a node needs a route, it floods ROUTE_REQUEST across the network. The destination replies with ROUTE_REPLY along the reverse path. |
| Reactive Routing | trades | Low Overhead vs. Discovery Latency | No periodic updates (saves bandwidth for sparse traffic), but incurs 200-500ms discovery delay when first packet to a destination is sent. |
| Reactive Routing | suited for | Sparse/Bursty Traffic | When nodes communicate infrequently (<1 packet every 5 minutes), on-demand discovery overhead is less than maintaining unused routes proactively. |
| Hybrid Routing (ZRP) | combines | Proactive (Intra-Zone) + Reactive (Inter-Zone) | Maintains proactive routes within local zones (e.g., 2-hop radius) for low-latency local delivery, and uses reactive bordercast for distant destinations to reduce overhead. |
| Hybrid Routing | suited for | Mixed Traffic Patterns | Balances proactive and reactive trade-offs when some destinations are frequently contacted (local zone) while others are rare (inter-zone). |
| Routing Overhead | scales as | O(n²) for Proactive | In flat proactive protocols, n nodes each maintain n-1 routes, resulting in n×(n-1) total routing state and periodic updates that grow quadratically with network size. |
| Routing Overhead | scales as | O(traffic) for Reactive | Reactive overhead is proportional to the number of route discoveries, which depends on traffic volume and route cache timeouts, not network size. |
| Route Cache | reduces | Discovery Frequency in Reactive Protocols | DSR caches discovered routes for a timeout period (e.g., 300 sec). Subsequent packets to the same destination reuse the cached path, avoiding redundant floods. |
| Route Cache | becomes stale | In Mobile Networks | Node mobility invalidates cached routes. If cache timeout exceeds link lifetime, using stale routes triggers ROUTE_ERROR messages and rediscoveries, negating cache benefits. |
| Sequence Numbers (DSDV) | prevent | Routing Loops | Each route advertisement includes a monotonically increasing sequence number. Fresher routes (higher seq#) always win, ensuring stale information doesn’t create forwarding loops. |
| Flat Routing | limits | Scalability to ~100-200 Nodes | All nodes are peers, requiring each to store and update routes to all destinations. Beyond 100-200 nodes, routing overhead dominates bandwidth. |
| Hierarchical Routing | improves | Scalability via Clustering | Grouping nodes into clusters with designated cluster heads reduces routing state from O(n) per node to O(k) (cluster size), enabling networks of 500+ nodes. |
| Geographic Routing (GPSR) | eliminates | Routing Tables | Forwarding decisions use GPS coordinates (greedy forwarding to neighbor closest to destination), requiring zero routing state or discoveries. |
| Geographic Routing | depends on | GPS Availability | Requires nodes to have location information (GPS, indoor positioning). Fails in GPS-denied environments (tunnels, buildings) where topology-based routing is needed. |
| Geographic Routing | immune to | Topology Changes | Mobility doesn’t invalidate routes because forwarding uses current GPS positions (updated every few seconds), not cached topology. Link breaks automatically handled by next beacon. |
| Link Break | triggers | Route Invalidation | Detected via missing ACKs (MAC layer) or missing beacons. Broken routes marked invalid (hop count = ∞) and ROUTE_ERROR sent to source (reactive) or propagated network-wide (proactive). |
| Mobility | challenges | Proactive Routing | Node movement causes frequent topology changes, triggering continuous routing updates. In high-mobility scenarios, triggered updates can exceed periodic update bandwidth. |
| Mobility | challenges | Reactive Routing | Cached routes become stale quickly. Frequent link breaks trigger ROUTE_ERROR storms and rediscoveries, potentially exceeding the overhead of proactive updates. |
| Friis Equation | explains | Multi-Hop Energy Efficiency | Received power ∝ 1/distance². Doubling distance requires 4× transmit power. Splitting long hop into short hops reduces source node’s energy consumption by orders of magnitude. |
| Broadcast Traffic | doesn’t require | Multi-Hop Routing | Single-hop broadcasts (e.g., beacon messages, status updates to neighbors) use MAC-layer broadcast, not network-layer routing. Only unicast multi-destination traffic needs routing protocols. |
Key Insight from Relationships:
The three routing paradigms form a trade-off triangle:
Low Latency
(Proactive)
/ \
/ \
/ \
/ \
Low Overhead -------------- Mobility Resilience
(Reactive) (Geographic/Hybrid)No protocol wins on all dimensions - network designers must prioritize based on application requirements (latency-critical? sparse traffic? GPS available?) and operational constraints (bandwidth budget, battery life, node mobility).
Protocol Selection Decision Tree:
Expand your understanding of multi-hop routing by exploring related protocols, wireless technologies, and system design topics:
Routing Protocol Deep Dives:
Foundation Concepts:
Related Routing Approaches:
Wireless Technologies:
System Design Considerations:
Real-World Applications:
Advanced Topics:
| If you want to… | Read this |
|---|---|
| Study proactive DSDV in depth | DSDV Proactive Routing |
| Learn reactive DSR routing | DSR Fundamentals and Route Discovery |
| Explore hybrid ZRP approach | Ad Hoc Routing: Hybrid (ZRP) |
| Learn about DTN for disrupted networks | DTN Store-Carry-Forward |
| Review ad hoc network applications | Ad-Hoc Network Applications |