Computer Engineering is coming soon

We're still writing and reviewing the notes for every subject, and we'll open it when every subject is ready. Meanwhile you can practise aptitude, ask Aria about any subject, or see the branches that are ready.

IoT: definition, characteristics and challenges

Definitions, core properties and main design challenges of the Internet of Things (IoT), with exact definitions, step-by-step derivations, formulas, worked numerical examples and common exam mistakes.

Drafted with Aria, reviewed by the AiCanCode.org team · Updated . How we write and check our notes. Spotted an error? Use Give Feedback at the bottom of the page.

Why it matters

IoT (Internet of Things) connects physical devices and sensors to the network so engineers can collect data, control devices and build automated systems for smart cities, industry and health. Understanding the definition, characteristics and challenges helps engineers choose sensors, networking, power management and security methods when designing an IoT system or answering exam questions about system trade-offs.

Key ideas

  • Thing / Device / Node — a hardware unit that contains one or more sensors (to measure physical properties), optional actuators (to perform actions), local processing (microcontroller), power source, and a communication interface. Draw: a small block labelled "Node" with inside boxes: Sensor(s), MCU, Radio, Power. Arrows: Sensor -> MCU, MCU -> Radio, MCU -> Actuator.

  • Sensor — a component that converts a physical quantity (temperature, light, acceleration) to an electrical signal.

  • Actuator — a component that converts an electrical command into physical action (motor, relay, valve).

  • Gateway — a device that bridges local IoT networks (often low-power, local protocols) to the Internet or cloud (IP networks). Draw: a local network of nodes connecting to a Gateway which connects to Internet/Cloud.

  • Cloud — remote servers and services that store, process and present IoT data; cloud provides long-term storage, analytics and user interfaces.

  • Latency — the time delay between an event (e.g., sensor data generation) and the time it is available at its destination (e.g., cloud or controller). Measured in seconds (s) or milliseconds (ms).

  • Data rate — the maximum amount of data that can be transmitted per unit time on a link, measured in bits per second (bit/s). Note: in communications theory, bandwidth B refers to frequency width and is measured in hertz (Hz); do not confuse data rate (bit/s) with bandwidth in Hz.

  • Throughput — the actual data successfully delivered per second. Distinguish from bandwidth: throughput ≤ bandwidth.

  • Duty cycle — fraction of time a device or radio is active (transmitting or receiving) over the total period. Important for energy calculations; define precisely: duty cycle = active_time / total_period.

  • Energy and Power — Power (P) is the rate of energy usage (watts, W). Energy (E) is power integrated over time (joules, J). Use E = P·t for constant power.

  • Availability — fraction of time a system is operational: availability = uptime / (uptime + downtime). Expressed as decimal or percentage.

  • Scalability — the property that a system continues to work acceptably when the number of devices increases. Discuss network addressing (IPv6), routing and data volume as scaling problems.

  • Interoperability — ability of heterogeneous devices and protocols to work together. Includes data formats, APIs and protocol compatibility.

  • Heterogeneity — variety in devices (different CPU, power, OS), protocols, and data formats; drives design complexity.

  • Security — protection against unauthorised access and attacks (confidentiality, integrity, availability). Define the three CIA pillars: Confidentiality (only authorised access), Integrity (data not altered), Availability (services accessible when needed).

  • Privacy — protecting personal information of individuals from inappropriate use or exposure.

  • Reliability — the probability a device will perform correctly over time; often related to mean time between failures (MTBF) and mean time to repair (MTTR).

  • Common exam diagram: "Three-layer architecture". Draw three horizontal layers top-to-bottom: Application layer (apps and users), Network layer (gateways, Internet), Perception layer (sensors/actuators). Label arrows up/down.

Derivations often required in exams:

  • Battery life with duty cycle: derive average power and operating hours from currents in active and sleep states. Step-by-step derivation will appear in worked examples.
  • Transmission time and energy: derive tx_time = bits / bitrate and energy = V·I·t = P·t.

Assumptions to state when answering theory or numerical: ideal radio (no retransmissions), fixed packet sizes, constant voltage supply, single-hop or known multi-hop path, AWGN channel if invoking Shannon capacity.

Formulas

E = P · t — E is energy (joule, J), P is power (watt, W), t is time (second, s). Applies when power is constant. P = V · I — P is power (W), V is voltage (volt, V), I is current (ampere, A). Applies to DC circuits. Battery_life_h = Capacity_mAh / Average_current_mA — Battery_life_h in hours (h); Capacity_mAh in milliampere-hour (mAh); Average_current_mA in mA. Applies for approximate battery life with roughly constant average current. Duty_cycle = t_active / T_period — Duty_cycle is unitless (fraction); t_active is active time (s); T_period is total period (s). tx_time = (Payload_bytes + Overhead_bytes) · 8 / Bitrate — tx_time in seconds; bytes to bits multiply by 8; Bitrate in bit/s. Availability = Uptime / (Uptime + Downtime) — unitless fraction or percentage. For repairable systems in steady state, an equivalent form is Availability ≈ MTBF / (MTBF + MTTR). Throughput_efficiency = Payload_bytes / (Payload_bytes + Overhead_bytes) — unitless fraction. C = B · log2(1 + S/N) — Shannon-Hartley channel capacity C in bit/s, B is bandwidth in Hz, S/N is signal-to-noise power ratio (linear). Applies to an ideal AWGN channel with optimal coding. Energy_per_bit = P / Data_rate — Energy per bit in joule/bit if P is average power (W) and Data_rate is bit/s.

Worked examples

  1. Battery life for a periodic sensor node (numerical). Problem: A sensor node wakes every 10 minutes to measure and transmit data. Transmission lasts 0.5 s and draws 40 mA at 3.3 V. Sleep current is 10 μA. Battery capacity is 2000 mAh (nominal 3.7 V). Estimate battery life in hours assuming only currents matter (ignore MCU processing energy) and duty cycling. Show steps. Step 1: Convert times: period T = 10 min = 600 s. Active time t_active = 0.5 s. Duty cycle = 0.5 / 600 = 0.0008333. Step 2: Average current I_avg = I_tx·duty + I_sleep·(1 - duty). I_tx = 40 mA, I_sleep = 10 μA = 0.01 mA. I_avg = 40·0.0008333 + 0.01·(0.9991667) = 0.033333 + 0.0099917 ≈ 0.043325 mA. Step 3: Battery life hours = Capacity_mAh / I_avg_mA = 2000 / 0.043325 ≈ 46,163 h. Step 4: Convert to years: 46,188 h / 8760 ≈ 5.27 years. Final answer: ≈ 5.3 years. (Assumptions: battery capacity usable fully, no self-discharge, constant current draw during tx, ignore voltage difference between battery and radio; state these in exam answers.)

  2. Transmission time and energy (numerical). Problem: Transmit 100-byte payload using IEEE 802.15.4 physical rate 250 kb/s. MAC+PHY overhead is 25 bytes. Radio TX current is 40 mA at 3.3 V. Compute tx time and energy for one packet. Step 1: Total bytes = 100 + 25 = 125 bytes. Bits = 125·8 = 1000 bits. Step 2: Bitrate = 250,000 bit/s. tx_time = 1000 / 250000 = 0.004 s = 4 ms. Step 3: Power P = V·I = 3.3 V · 0.04 A = 0.132 W. Step 4: Energy E = P·t = 0.132 · 0.004 = 0.000528 J = 528 μJ. Final answers: tx_time = 4 ms, energy = 528 μJ. (Assumptions: no retransmissions, preamble/turnaround overhead included in 25 bytes; if not, mention additions.)

  3. Short theory: Describe three major IoT challenges and how engineers mitigate them (model answer).

  • Power management: devices are often battery-powered and must last months/years. Mitigations: aggressive duty cycling, use of low-power radios, energy-aware MAC protocols, energy harvesting when possible, and local processing to reduce transmissions (edge filtering). Mention the battery life derivation step.
  • Security and privacy: many devices have weak compute power for strong crypto, but data may be sensitive. Mitigations: lightweight cryptographic primitives, secure boot, device identity management, TLS/DTLS gateways, and secure firmware update mechanisms.
  • Interoperability and heterogeneity: devices use different protocols and data formats. Mitigations: gateways, common data models (refer to later topic ‘IoT ecosystem’ and APIs), use of standards (MQTT, CoAP) and mapping layers. (Exam answers should define each challenge, explain why it is a problem, and list concrete mitigation techniques.)

Common mistakes

  • Forgetting to convert units (bytes→bits or ms→s) in transmission time calculations.
  • Using battery capacity in mAh directly with power in watts without converting or using average current.
  • Not stating assumptions (no retransmissions, perfect channel, constant voltage) in numerical answers.
  • Mixing availability and reliability definitions.
  • Omitting overhead bytes (MAC/PHY headers) when calculating transmission time or throughput efficiency.

For your exam

  • Short notes: Explain characteristics (scalability, heterogeneity, energy-constrained) with 2–4 mitigation points each.
  • Derivations: Derive battery life with duty cycle, or derive average power: P_avg = P_active·duty + P_sleep·(1 - duty). Show all steps and unit consistency.
  • Numericals: Typical problems on battery life, tx time and energy, throughput efficiency. Show conversions (bytes→bits), list overhead, and give final answer with units.
  • Diagrams: Draw and label the three-layer architecture or a node block diagram. A full-marks diagram is neat, labelled, and referenced in the explanation.
  • Long question (theory): Compare protocols or discuss design trade-offs: include definitions, constraints, trade-offs and specific mitigation strategies.

A full-marks answer always: (1) states definitions and assumptions, (2) shows step-by-step derivation/calculation with units, (3) comments on limitations/assumptions, and (4) where appropriate, draws and labels a simple diagram.

Quick check

  1. Define duty cycle and give its unit.
  2. Compute tx_time for a 50-byte payload with 10 bytes overhead at 125 kb/s.
  3. What is the formula relating energy, power and time? Answers: 1) Duty cycle = t_active / T_period, unitless fraction (or percent). 2) Total bits = (50+10)·8 = 480 bits; tx_time = 480 / 125000 = 0.00384 s = 3.84 ms. 3) E = P · t.

Finished this topic? Mark it so your progress, study plan and readiness keep up.

Stuck on something here?