IoT Communication Protocols: A Comprehensive Guide
IoT Communication Protocols
ZigBee:
- Features: 2.4GHz ISM band, low power consumption, mesh networking capability
- Advantages: Simple protocols designed for low power/cost devices, self-healing mesh extends range
- Disadvantages: Lower data rates (250kbps), shorter range compared to other standards
Wi-Fi HaLow (802.11ah):
- Features: Operates in sub-GHz frequencies (900MHz), designed for longer range
- Advantages: Extended range up to 1km, good penetration through obstacles
- Disadvantages: Potential interference from other sub-GHz devices, adoption still ongoing
Wi-Fi 6 (802.11ax):
- Features: 2.4/5GHz bands, OFDMA, target wake time scheduling, 20MHz-only operation
- Advantages: Higher efficiency/throughput, better battery life from scheduling
- Disadvantages: Requires new Wi-Fi 6 capable hardware to take full advantage
Sigfox:
- Features: Ultra narrowband signals, star network topology, 3 redundant uplink channels, pseudo-random frequency hopping
- Advantages: Very long range (~40km), extremely low power consumption
- Disadvantages: Extremely low data rates
LoRaWAN:
- Features: Sub-GHz unlicensed spectrum, long range, low power, low bitrate
- Advantages: Open standard, good range (2-5km urban), low power allows battery operation
- Disadvantages: Lower data rates around 50kbps, higher latencies
NB-IoT:
- Features: 3GPP Rel 13 LTE evolution, direct to 4G integration, power saving modes
- Advantages: Leverages existing cellular infrastructure, good range, lower power
- Disadvantages: Requires subscription fees, complexity of cellular integration
BLE:
- Features: Bluetooth Low Energy, ad-hoc piconets/scatternets
- Advantages: Extremely low power consumption, simple protocols
- Disadvantages: Limited range around 100m, interference issues
ANT+:
- Features: Low power wireless protocol tailored for sensors
- Advantages: Ultra-low power, simple stack ideal for fitness/health tracking
- Disadvantages: Very short range around 30m, narrow use case
IoT Communication Protocol Selection
Choice depends on factors like:
- Frequency of communication (event-based or sampling rate)
- Uplink/downlink requirements
- Range
- Power budget
- Interoperability needs
- Reliability requirements
Need to consider the transmission impairments like attenuation (signal energy loss) which requires amplifiers to compensate.
Transmission Impairments
- Attenuation: The gradual loss of signal strength or power as the signal travels through a medium.
- Distortion: Refers to the alteration or deformation of the transmitted signal’s waveform.
- Noise: Refers to any unwanted signals or interference that corrupts the transmitted data signal.
Design Goals for IoT
- Low power consumption
- Lightweight in memory
- Interoperability
Tasks of an Operating System (OS) in IoT
- Power/Memory management
- Process scheduling
- Command interpreter
- User interface
Operational Definitions
- Reliability: Uptime of service
- Survivability: How long to recover (intermittent or permanent)
- Availability: Not available for the service it is designed for
Network Lifetime
- First node failure
- Last node failure
- Failure of the first critical component that can cause network partition (also known as a pivotal node)
- Until the network permanently fails to deliver all its functional requirements.
Design Considerations for Medium Contention Plan
- Performance Bottlenecks: Delays, packet loss (performance metrics), type of protocol (TDMA, etc.), event-based vs. sampling rate, self-healing, power considerations
Buffer Datagram
It allows near queue if data is pre-ready in the buffer. If buffering capacity is packed, packets can be dropped due to buffer overflow or priority competition.
Buffer Overflow
If arrival rate > processing rate, it causes a bottleneck and buffer overflow.
Increasing buffer size is not a good solution because:
- Wait time for the buffer increases (infinitely if the buffer size is infinite).
Solutions for Buffer Overflow
- Increase the number of devices (servers)
Network Neutrality and Congestion Control
Network neutrality means all traffic goes through the same fate. Congestion control mechanisms help manage network traffic to prevent bottlenecks.
Packet Dropping Policies
When the buffer is full, decisions need to be made about which packet to drop. Common policies include:
- Drop Tail: Drop the last packet (if using FCFS)
- Other Policies: Weighted fair queuing, overwriting already queued packets
Buffer Sizing
How much buffering is required?
RTT = Round Trip Time; c = Link Capacity
Handling Out-port Queues
- FIFO (First-In, First-Out): Simple, fair (no priority)
- Packet Scheduler: Can implement various queuing disciplines like FCFS, weighted fair queuing, etc.
Active Queue Management (AQM)
- Drop packets before the buffer is full to signal congestion early.
- Congestion control mechanisms can then throttle traffic to avoid buffer overflow.
- Random Early Detection (RED):
- Set two thresholds (MINthrsh & MAXthrsh).
- If Qavg > MINthrsh, start dropping packets with a probability (Pdrop) proportional to Qavg.
- If Qavg > MAXthrsh, drop all incoming packets.
- This helps control the sampling rate.
Head-of-the-Line (HOL) Blocking
Occurs when a queued packet using the medium blocks subsequent packets in the queue from moving forward.
Network Topologies
- Infrastructure Mode (Hierarchy): Centralized control, typically with an access point.
- Handoff Mode: Devices communicate directly with each other, forming a mesh network.
