The architecture of sensor networks, specifically wireless sensor networks (WSNs), is tailored to

accommodate the specific needs of applications and the communication technologies they rely on.
This results in significant variations in both individual sensor node hardware and the overall network
structure.

1. Sensor Node Hardware Components:

- Microcontrollers:

- Sensor nodes are built using simple embedded microcontrollers (e.g., Atmel or Texas Instruments
MSP430). These microcontrollers must balance power consumption with computational power,
memory, and sleep modes, which influence energy efficiency and responsiveness.

- Radio Transceivers:

- Common transceivers include RFM TR1001 and Chipcon devices, which use modulation techniques
like ASK, FSK, or OOK (as in Berkeley PicoNodes). Emerging technologies, such as ultra-wideband,
offer advances in communication efficiency. A potential breakthrough is the wake-up radio concept,
which allows sensor nodes (SNs) to remain in sleep mode and wake up when a relevant transmission
is detected, reducing power consumption.
- Energy Supply (Batteries):

- Batteries are the primary power source, but efficient battery management and energy scavenging
(recharging through environmental energy) are critical for long-term operation. Self-discharge and
recharge rates, as well as overall battery life, are key concerns.

- Operating Systems & Run-Time Environments:

- The trade-off between minimal memory and flexibility for protocol integration is central to the
design of operating systems for SNs. Run-time environments that support structures like
blackboards, publish/subscribe, or tuplespaces are suggested as suitable for handling the multi-
protocol, meta-information-heavy requirements of WSNs.

2. Network Architecture:
The architecture of WSNs must ensure both sensing coverage and communication connectivity,
which relies on the density of the sensor deployment.

- Density and Lifetime:

- The network's lifetime can be defined by either the failure of individual sensors or by a percentage
failure rate (e.g., 95% area coverage). Denser sensor deployment provides redundancy, but
increasing density can also lead to higher energy consumption and transmission delays due to
collisions. The optimal sensor density maximizes sensor lifetime by balancing communication
distance and energy consumption.

- Transmission Strategies:
- Energy consumption varies depending on whether data is transmitted directly between two
sensors or via intermediate nodes. Using intermediate sensors can save energy compared to long-
range transmissions (Figure 9.3 illustrates these strategies).

- Energy-Driven Protocols:
- WSN protocols must be energy-efficient, focusing on data aggregation and reducing unnecessary
transmissions. Redundant or correlated sensor readings can be aggregated to minimize the number
of packets sent.

- Addressing and QoS:

- Address-free structures may be essential for scalability and energy efficiency, using geographic or
data-centric addressing systems. QoS, redundancy, and sensor reading imprecision must be
accounted for in the design of the network.
In-Network Processing:
- WSNs often perform in-network processing, such as data aggregation and distributed signal
processing, to reduce the number of transmitted packets. The exact nature of WSN services is not
fully defined, as they do not simply transmit bits like traditional networks.

- Gateway and Integration:

- WSNs must be integrated into larger networks via a gateway to bridge different communication
protocols. Integration into middleware architectures, like CORBA or Web Services, and services
described in WSDL or UDDI present challenges in representing the unique functionality of WSNs.

- Reconfiguring Tasks:
- It may be necessary to update the tasks or software of sensor nodes over time, requiring remote
reconfiguration capabilities for the WSN.

Physical Layer
The physical layer in the context of wireless sensor networks (WSNs) refers to the hardware and
communication technologies used to transmit data between sensor nodes. However, limited
research has been conducted on developing protocols specifically tailored to the unique needs of
WSNs at the physical layer, particularly in terms of energy efficiency and handling the overhead
related to radio transmission (e.g., retransmissions and communication costs).

Several points about the physical layer include:

- Energy-Efficient Transmission: A key focus is on how to transmit data with minimal energy
consumption while considering potential retransmissions and overhead. Some energy-efficient
modulation techniques have been explored in previous research (e.g., Schurgers 2001).

- Modulation Techniques for Sensor Nodes: Research such as that by Gao (2001) addresses the
hardware aspects of CDMA (Code Division Multiple Access) and other modulation issues,
emphasizing how these impact sensor nodes' performance.
- Communication Protocol Design: Shih (2001b) provides discussions on communication protocol
designs that are based on the physical layer, offering insights into how the physical layer impacts
overall network performance.

- Standardization Efforts: Although specific details are not covered, the work being done by
organizations like the IEEE (notably the IEEE 802.15.4 standard) is recognized as relevant. IEEE
802.15.4 focuses on low-rate wireless personal area networks (LR-WPANs), which are critical for
many WSN applications due to their low power consumption and cost-effective communication.

MAC LAYER
The MAC (Medium Access Control) layer plays a vital role in wireless sensor networks (WSNs), mainly
in regulating how sensor nodes share the communication medium efficiently while minimizing
energy consumption. The main focus of MAC layer research for WSNs is to keep sensor nodes in
sleep mode as long as possible to conserve energy, making it a crucial area of study.

Key concepts of MAC layer include:

Channel Allocation Schemes:

- Static Channel Allocation: Bandwidth is divided among sensor nodes using techniques like FDMA
(Frequency Division Multiple Access), TDMA (Time Division Multiple Access), CDMA (Code Division
Multiple Access), or SDMA (Space Division Multiple Access). These methods minimize interference
but work best in static, low-node-count environments where nodes have predictable data loads.

- Dynamic Channel Allocation: This is more suited for environments where the number of active
nodes varies, and data generation is bursty (i.e., sporadic). Contention-based protocols such as CSMA
(Carrier Sense Multiple Access) are used here. Although these can minimize collisions, they may lead
to energy wastage and delays due to retransmissions.

Hierarchical Clustering and TDMA:

- In a hierarchical clustering model, TDMA can be used to allocate time slots to each sensor node
within a cluster. Nodes transmit data in their designated slots and can sleep during other periods to
save energy. This is especially beneficial in both proactive networks (where nodes transmit
periodically) and reactive networks (where nodes respond to sudden events).
CDMA and Intra/Inter-Cluster Collisions:

- CDMA can be used to avoid collisions between clusters. Though more energy is required per bit
transmitted, CDMA allows multiple transmissions using the same frequency, making it useful for
inter-cluster communication.

MAC Design Challenges:

1. Node Failure: When sensor nodes fail, MAC protocols must adapt by forming new routes and
adjusting power levels to maintain network connectivity.

2. Energy Consumption Sources:

- Collisions: Corrupted packets due to collisions lead to retransmissions, consuming more energy.

- Overhearing: Nodes may waste energy by listening to transmissions not intended for them.

- Control Packet Overhead: The need for control packets to manage network operations increases
energy usage as the number of nodes rises or nodes fail.

- Idle Listening: Nodes may waste energy waiting to receive data that is never sent.

Energy Conservation Techniques:

- One of the most effective methods for conserving energy in WSNs is periodic sleeping: nodes turn
off their radios and synchronize based on a sleep schedule. This reduces the idle listening problem, a
major energy drain in sensor networks.

Scheduling and Reservation vs. Contention-based MAC:

- Scheduling-based MAC protocols (like TDMA) save more energy compared to contention-based
protocols (like IEEE 802.11). However, TDMA requires careful management of cluster formation,
inter-cluster communication, and dynamic adaptation to changes in the number of nodes in the
network.

In summary, the MAC layer in WSNs faces unique challenges related to energy efficiency, node
failure, and dynamic network environments. Various strategies, including static and dynamic channel
allocation, hierarchical clustering, and scheduled sleeping, are employed to balance energy
conservation with reliable communication.
The Sensor-MAC (S-MAC) protocol:

1. Energy Efficiency in WSNs:

- WSNs are designed for prolonged operation, making energy conservation crucial since recharging
batteries is often impractical.

- Nodes typically spend much of their time in an idle state, consuming similar energy levels in both
idle and receiving modes.

2. Low Duty Cycle Operation:

- Nodes in S-MAC operate on a low duty cycle, alternating between listening and sleeping to
conserve energy.

- By doing this, S-MAC addresses four primary sources of energy consumption: collision, control
overhead, overhearing, and idle listening.

3. Coordinated Sleep Schedules:

- S-MAC nodes communicate and coordinate their sleep schedules to minimize energy waste.

- Each node broadcasts its schedule to neighbors using SYNC packets, helping to prevent clock drift
and ensuring nodes follow similar wake and sleep patterns.

4. Handling Different Schedules:

- If two neighboring nodes have different schedules, they can either:

- Listen to both schedules, or

- Transmit data according to both schedules.

- Neighbor discovery is periodically performed to ensure nodes are aware of each other's presence
and schedules.

5. Communication Process:

- During communication, a node first performs carrier sensing during its neighbor's listening
periods.

- If the channel is clear, it sends a Request to Send (RTS) followed by a Clear to Send (CTS) from the
receiver, enabling data transfer during their scheduled times.

- Broadcasts do not use RTS/CTS to avoid potential collisions from multiple responses.
6. Adaptive Listening:

- To reduce latency in multi-hop transmissions, nodes that overhear transmissions can briefly wake
up to receive data, rather than waiting for their next scheduled listen time.
 EAR protocol (Energy and Activity Aware Routing)
It is designed to facilitate communication between mobile and stationary nodes in a wireless sensor
network (WSN), particularly where energy efficiency and bandwidth are concerns. Here's a
breakdown of the key concepts:

1. Introduction of Mobility: EAR accommodates mobile nodes that can move through a network
primarily composed of stationary sensors. This is particularly useful in scenarios where limited power
consumption is crucial.

2. Low-Power Operation: The protocol is designed to minimize energy usage by allowing mobile
nodes to initiate and control connections with stationary nodes. This reduces the number of
messages exchanged, which is important since battery life is a priority for stationary nodes.

3. Connection Control: Mobile nodes manage when to connect or disconnect from stationary nodes.
They have a registry that tracks nearby stationary nodes, allowing them to make informed decisions
about connections based on proximity and signal quality.

4. Eavesdropping Mechanism: Mobile nodes listen for periodic "Broadcast Invite" messages sent by
stationary nodes. These messages inform them of available connections without requiring immediate
responses, allowing mobile nodes to assess the network environment.

5. Connection Establishment: The EAR protocol involves several key message types:

- Broadcast Invite (BI): Sent by stationary nodes to invite mobile nodes to connect. Mobile nodes
respond by registering the stationary nodes they encounter.

- Mobile Invite (MI): A response from the mobile node to a BI message, indicating its intent to
connect.

- Mobile Response (MR): Sent by the stationary node in reply to the MI, confirming the connection
and assigning communication slots in the TDMA frame.

- Mobile Disconnect (MD): Sent by the mobile node when it wants to terminate a connection,
typically based on the signal quality (SNR).

6. Efficiency and Transparency: The EAR protocol is designed to be transparent to the existing
stationary nodes' protocols, minimizing the need for specialized messages and ensuring seamless
operation within the network.
The Sparse Topology and Energy Management (STEM)
This protocol is designed for sensor networks that spend most of their time in a low-power
monitoring state, only becoming active when specific events occur. Here's a breakdown of its main
concepts:

1. Reactive Monitoring: STEM is particularly suited for applications where the network primarily
senses the environment, such as detecting forest fires. The goal is to keep the network operational
for long periods while using minimal energy during monitoring.

2. Energy-Efficient Operation: During the monitoring state, only the sensors and some preprocessing
circuits of the nodes are active. The main processor and communication radio remain off to save
energy until an event is detected.

3. Event Detection: When a potential event is detected, the node's main processor wakes up to
analyze the data and determine if communication with other nodes is necessary. However, the radios
of neighboring nodes may remain off if they did not detect the same event.

4. Beacon Communication: To initiate communication, a node (the initiator) periodically sends out a
beacon signal targeting another node (the target). This beacon alerts the target node to wake up and
establish a communication link.

5. Link Activation: Upon receiving the beacon, the target node responds, and both nodes keep their
radios on. If the data needs to be forwarded, the target becomes the initiator for the next hop,
repeating the process.

6. Separate Frequency Bands: STEM uses different frequency bands for the wakeup protocol and
data transmission to avoid interference. The wakeup messages are sent over one frequency (the
"wakeup plane"), while actual data packets are transmitted over another frequency (the "data
plane").

7. Hardware Support: Some commercially available sensors support this dual-band operation,
allowing for effective implementation of the STEM protocol.
 ROUTING LAYER
This section explores energy-efficient routing techniques developed specifically for WSNs,
focusing on extending network lifetime.
Key concepts of WSN routing include:
Routing Approaches:
1. Multi-Hop Routing: In WSNs, data is usually sent over multiple hops from a source
node to the BS. The BS can be fixed or mobile, serving as a gateway between the
sensor network and external systems, such as the Internet. Multi-hop routing is
energy-efficient, as direct communication with the BS can be energy-intensive for
distant nodes.
2. Energy-Efficient Routing Goals: The primary aim is to minimize energy consumption
and maximize network lifetime. WSN routing protocols use specific strategies,
including data aggregation, clustering, in-network processing, and data-centric
techniques to meet this goal.
 Classification of Routing Protocols for WSNs

Routing protocols in WSNs can be classified based on network structure and protocol
operation.
By Network Structure:
• Flat-based Routing: All nodes have equal roles, and data is typically forwarded based
on query responses. This approach is suitable for small networks.
• Hierarchical-based Routing: Nodes play distinct roles; some nodes (cluster heads)
are designated with additional responsibilities, such as aggregating data and
forwarding it to the BS. This approach conserves energy by reducing redundant data
Transmission and is commonly used in large WSNs.
• Adaptive Routing: Protocols adapt parameters based on current network conditions,
such as energy availability, to maintain efficiency.
By Protocol Operation:
• Multipath-based Routing: Utilizes multiple paths for data transmission to increase
reliability and load balancing.
• Query-based Routing: Nodes send data in response to specific queries, saving energy
by limiting transmissions to relevant data only.
• Negotiation-based Routing: Prevents redundant data transmission through
negotiation among nodes before data forwarding.
• Location-based Routing: Nodes use location information to route data, optimizing
routes and reducing transmission energy costs.
Considerations for WSN Routing:
1. Energy Conservation Priority: Energy efficiency is prioritized over data quality to
prolong network lifespan.
2. Stationary Nodes: Many applications assume nodes are stationary, favoring table-
driven routing protocols over reactive schemes to reduce energy spent on route
discovery.
3. Cooperative Routing: Sensor nodes collaborate by sending data to a cluster head
(CH), which aggregates the data and further processes it, reducing overall route cost
in terms of energy use.
4. Use of Timing and Position Information: Certain protocols leverage time and
position data for efficient routing, adapting routes as network conditions and node
locations change.

Network Structure Based

Flat Routing
In flat routing based protocols, all nodes play the same role.
Directed Diffusion
Directed Diffusion operates through a process involving data aggregation and in-network
processing, which helps reduce redundant transmissions and save energy. Here’s an
overview of the Directed Diffusion process:
1. Interest Propagation: The base station (BS) initiates communication by broadcasting
an "interest" that specifies a task to be fulfilled by the network. This interest diffuses
through the network hop-by-hop, with each node broadcasting it to its neighbors.
2. Gradient Setup: Each node that receives an interest establishes a gradient—a
direction of data flow—towards the node from which it received the interest. This
gradient setup forms paths from multiple data sources to the BS, allowing data to
flow towards the requesting node. Different gradients may be formed to neighbors
with varying strengths, affecting data flow.
3. Data Transmission and Path Reinforcement: Once data is transmitted, paths are
refined by reinforcing the best paths to ensure efficient routing and prevent further
flooding. The BS may also periodically refresh the interest to maintain data flow, as
the wireless medium can be unreliable.
4. In-network Aggregation: As data flows along these gradients, nodes aggregate
information from various sources, reducing the number of transmissions and
minimizing energy consumption.
 5. Event-Specific Propagation: Directed Diffusion is well-suited for ongoing or persistent
queries, where nodes continuously provide data over time, such as monitoring
environmental conditions. However, it may not be efficient for one-time or
historical queries, as setting up gradients for a single-use path can be unnecessary.
By aggregating data and choosing optimized paths, Directed Diffusion conserves energy,
extends network lifetime, and improves robustness.

Hierarchical Routing is a type of cluster-based routing approach in wireless

sensor networks (WSNs) aimed at improving energy efficiency and scalability. By organizing
nodes into clusters, hierarchical routing protocols achieve a balanced load across the
network, where energy-intensive tasks are primarily handled by selected nodes called
Cluster Heads (CHs).

Key Concepts of Hierarchical Routing

1. Cluster Formation and Role Assignment: Nodes in a WSN are organized into clusters, with
each cluster containing a designated CH. Regular sensor nodes within a cluster perform
sensing and data collection tasks and then relay this data to their CH. The CH, typically a
higher-energy node, is responsible for aggregating data and sending it either to an upper-
level CH or directly to the Base Station (BS).
2. Energy Efficiency: By aggregating data, CHs reduce the number of transmissions and the
energy spent in data transfer. Since data is aggregated at the cluster level before reaching
the BS, this hierarchical approach minimizes communication cost, conserves node energy,
and enhances the network's lifespan.

3. Scalability: Hierarchical routing supports a scalable architecture by limiting

communication within clusters and only forwarding processed data to higher layers. This
structure is especially beneficial in large networks where a direct transmission from each
node to the BS would be inefficient and unsustainable.

Example: Cluster-Based Routing Protocol (CBRP)

The Cluster-Based Routing Protocol (CBRP) is a fundamental protocol proposed by Jiang
(1998) that segments the network into clusters of two-hop-diameter. Each node only
communicates with its CH, which is responsible for further routing data. However, CBRP has
some notable limitations for WSNs:
- High Overhead: CBRP requires frequent "hello" messages to establish and maintain
clusters, which consumes significant network energy and bandwidth.
- Stationary Nodes: In most WSN applications where nodes are stationary, the overhead
created by continuous hello messaging becomes unnecessary and reduces overall efficiency.

Adaptive Routing protocols are designed to dynamically adjust to the available resources
and network environment, aiming to conserve energy and improve communication
efficiency in Wireless Sensor Networks (WSNs). One prominent family of adaptive routing
protocols is the
Sensor Protocols for Information via Negotiation (SPIN)
Key Concepts of Adaptive Routing with SPIN
1. Data-Centric Communication: SPIN protocols assign high-level metadata to data
packets, allowing nodes to negotiate the data before transmission. This data-centric
approach prevents the transmission of redundant information since nodes only send
data that their neighbors lack.
2. Energy Adaptation: SPIN adapts to each node’s available resources, ensuring that
low-energy nodes reduce participation when energy is scarce. This energy-awareness
prevents nodes from expending energy on unnecessary transmissions, prolonging
the network’s operational lifespan.
 3. Three-Stage Communication: SPIN protocols follow a three-stage communication
process using three types of messages:
o ADV: Advertises new data.
o REQ: Requests specific data that the node does not possess.
o DATA: Transmits the actual data when requested.
This process continues iteratively until all nodes in the network have the information they
need.
Types of SPIN Protocols
• SPIN-1 and SPIN-2: These are the foundational protocols of the SPIN family. SPIN-1
handles basic data negotiation, while SPIN-2 includes a threshold-based energy
mechanism. When a node's energy level drops below a certain point, it only
participates in data transmission if it can complete the process without falling below
the threshold.
• SPIN-BC (Broadcast Channel): SPIN-BC is optimized for broadcast environments.
Nodes delay their responses with random timers, avoiding redundant requests when
multiple nodes request the same data.
• SPIN-PP (Point-to-Point): For point-to-point communication, SPIN-PP operates with a
three-way handshake similar to SPIN-1 but requires only single-hop neighbor
awareness. This protocol is effective for networks where energy and data reliability
are less of a concern.
• SPIN-EC (Energy-Constrained): This variant uses an energy threshold heuristic. Nodes
participate only if they can complete all stages without reducing energy below a
predefined threshold.
• SPIN-RL (Reliable): In error-prone environments, SPIN-RL adds mechanisms to SPIN-
PP to manage lossy channels. Nodes track received ADV messages and can request
retransmissions for data not received within a time window, ensuring reliable data
transfer.
Advantages of SPIN
• Energy Efficiency: By negotiating metadata before full data transfer, SPIN reduces
unnecessary data exchanges and conserves energy.
• Adaptability: SPIN adjusts to the network’s energy levels and only participates when
feasible, extending node life.
• Reduced Redundancy: By eliminating redundant data transfers and only transmitting
unique data, SPIN conserves bandwidth and energy.
Limitations
SPIN protocols, while energy efficient and suitable for mobile WSNs, are not always ideal for
networks with strict latency requirements or real-time demands. The delay from
negotiation and the overhead from ADV and REQ messaging can be excessive in dense
networks or environments where rapid data dissemination is needed.