Wireless Communication¶

Exam-ready answer

The cellular concept divides a large service region into many low-power radio coverage areas called cells, each served by a BTS, eNodeB or gNodeB. A mobile uses an assigned uplink/downlink channel to the serving base station; base stations connect through controllers or an IP radio-access network to the switching/core network. As the user crosses a boundary, handover transfers the session to a neighbor. Capacity rises because the same channel set can be reused by sufficiently separated, nonadjacent cells rather than one high-power transmitter serving the whole region.

Real coverage contours are irregular because terrain, buildings, antenna patterns and fading alter received power. A hexagon is therefore a planning model, not the literal RF boundary. It is preferred because it tessellates a plane without gaps or overlaps, has six equal-distance adjacent cell centers, approximates a circle better than a square, and provides simple cluster/reuse geometry. Circles approximate equal-range propagation but cannot tile; triangles and squares tile but represent circular coverage less closely and have fewer equidistant neighbors.

For ideal hexagon side/cell radius \(R\), area is

when \(R\) is in km. Thus \(R=2\,\text{km}\) gives \(A_c\approx10.39\,\text{km}^2\). Macrocells provide wide rural coverage; micro-, pico- and femtocells improve urban/indoor capacity at lower power. Cellular planning supports frequency reuse, mobility and scalable capacity, but smaller cells require more sites, backhaul and handovers, while the ideal hexagon must be corrected using propagation measurements and interference analysis.

Practice target: 8 minutes; draw the mobile-to-core path, state four hexagon reasons and finish with one capacity benefit and one planning limitation.

5-mark answer and 10-mark extension

For 5 marks — reuse fundamentals¶

Frequency reuse assigns the total available channel set once within a cluster of \(N\) cells, then repeats the set in other clusters far enough away to keep interference acceptable. The reuse factor is \(1/N\); if the system has \(S\) duplex channels, an ideal cell receives about \(S/N\) channels. For a hexagonal layout,

where nonnegative integers \(i,j\) specify the cluster shift, \(R\) is cell radius and \(D\) is nearest co-channel center separation, both in the same length unit.

Cells assigned the same channel group are co-channel cells; their unwanted received energy is co-channel interference (CCI). It cannot be removed by a sharper channel filter because desired and interfering signals occupy the same frequency. Increasing \(N\) and \(D\), lowering unnecessary transmit power, sectoring, directional antennas, beamforming and interference coordination reduce CCI. A small \(N\) gives more channels per cell but worse CCI; a large \(N\) improves quality but reduces spectral reuse and capacity.

Add for a 10-mark reuse analysis¶

The allowed cluster sizes follow the geometry of moving \(i\) cell centers along one hexagonal chain, turning \(60^\circ\), then moving \(j\) centers. Repeating this translation gives \(N=i^2+ij+j^2\) and, from the center geometry, \(D=R\sqrt{3N}\). For \(i=2,j=1\), \(N=7\). If \(R=2\,\text{km}\),

For equal-power first-tier interferers and path-loss exponent \(n_p\), a useful edge approximation is

where \(i_0\) is the number of dominant co-channel interferers, commonly six for an omnidirectional first tier. With \(N=7\), \(n_p=4\) and \(i_0=6\), \(C/I\approx73.5\), or \(10\log_{10}(73.5)=18.7\,\text{dB}\). The approximation must be refined for unequal ranges, shadowing, traffic loading, antenna patterns and multiple tiers.

Adjacent-channel interference is different: energy from nearby frequency channels leaks through imperfect filters or near-far conditions, so guard bands, filtering and channel assignment help. CCI instead drives reuse planning. Cell splitting increases reuse density but adds sites and handovers; sectoring reduces the number of strong interferers but needs directional antennas and reduces trunking flexibility. Modern LTE/5G systems replace rigid frequency groups with reuse-one scheduling, power control, beamforming and inter-cell coordination, yet the same capacity-versus-interference principle remains.

Practice target: 9 minutes for definitions and geometry or 18 minutes with the \(C/I\) estimate, numerical reuse distance and mitigation trade-offs.

Exam-ready answer

Handover (handoff) transfers an ongoing voice or data bearer from one cell, sector, channel, beam or radio access network to another while preserving service. It is needed when a mobile leaves coverage, serving RSS/RSRP or SINR falls, interference or BER rises, a cell is overloaded, or another RAT offers better service. The mobile and base station measure the serving and neighbor cells; the network selects a target, reserves resources, transfers context, commands the mobile, switches the user path through the core, and releases the old resources.

To avoid ping-pong, a typical event requires

continuously for a time-to-trigger, often with a minimum serving threshold. If serving power is \(-92\,\text{dBm}\), neighbor power \(-86\,\text{dBm}\) and hysteresis \(H=3\,\text{dB}\), the neighbor leads by \(6\,\text{dB}\) and qualifies after the timer; transient crossings do not trigger.

GSM uses mobile-assisted measurements with a network decision; LTE/5G normally use hard handover, while 5G also manages beams and LTE-NR interworking. Correct hysteresis balances late handover/drop risk against early handover and ping-pong. Handover enables continuous mobility and load balancing but consumes signaling, may briefly interrupt data, and can fail if target resources, backhaul or radio conditions are inadequate.

Practice target: 8–9 minutes; draw the signal crossover, state the hysteresis/timer rule and compare hard, soft, horizontal and vertical cases.

5-mark answer and 10-mark extension

For 5 marks — generation summary¶

Mobile generations are major standards transitions in radio access, core-network operation and supported service. Their common physical path is user equipment → cellular base station/RAN → switching or packet core → external voice/data network.

Thus evolution is analog voice → secure digital voice/text → packet multimedia → all-IP broadband → software-defined, beamformed and service-tailored connectivity.

Add for a 10-mark generation analysis¶

FDMA separates users by guarded frequency channels; TDMA gives users recurring time slots on a carrier; CDMA lets users share time and frequency using distinct spreading codes and power control; OFDMA schedules orthogonal subcarrier-time resource blocks among users. These distinctions explain capacity growth: 1G allocates one analog frequency channel, GSM combines carrier and slot, CDMA exploits code-domain reuse, and LTE/NR schedule fine-grained orthogonal resources. NR can use CP-OFDM on both links and DFT-s-OFDM on uplink, with flexible subcarrier spacing; massive MIMO adds spatial separation.

The theoretical AWGN rate test remains

with bandwidth \(B\) in Hz and linear \(S/N\). For \(B=20\,\text{MHz}\) and \(S/N=10\,\text{dB}=10\), \(C=20\times10^6\log_2(11)\approx69.2\,\text{Mbit/s}\) before protocol, fading and interference losses. Later generations approach useful fractions of this limit through higher-order modulation, coding, wider aggregated bandwidth and MIMO rather than by generation name alone.

1G/2G wide-area voice emphasized coverage and battery life; 3G introduced practical mobile data but had complex CDMA planning; 4G flattened the packet architecture but still depends on spectrum and backhaul; 5G adds dense cells, edge functions and network slicing, with eMBB for high throughput, URLLC for bounded low-latency reliability, and mMTC for massive low-power IoT. Limitations include higher band blockage and rain/foliage loss, dense-site cost, device compatibility, energy use and no universal guarantee of headline rate or latency. Generations coexist through roaming, fallback and inter-RAT handover.

Practice target: 9 minutes for the generation table or 18 minutes with resource diagrams, the capacity check and deployment trade-offs.

Exam-ready answer

A wireless channel is the time-varying propagation path between transmitting and receiving antennas. Its large-scale mean falls with distance (path loss), obstacles cause slow log-normal shadowing, and replicas arriving with different delays, phases and angles cause small-scale multipath fading.

• Free space: a clear line-of-sight path follows Friis spreading, with $\(L_{fs}=32.44+20\log_{10}d_{km}+20\log_{10}f_{MHz}\ \text{dB}.\)$
• Two-ray ground model: direct and ground-reflected rays combine; beyond the crossover region the mean received power is approximately \(P_r=P_tG_tG_rh_t^2h_r^2/d^4\) W, so it decays roughly as \(d^{-4}\) rather than free-space \(d^{-2}\).
• Reflection occurs from surfaces large relative to wavelength; diffraction bends energy around edges and supplies some shadow-region coverage; scattering from rough surfaces and small objects spreads energy in many directions.

Multipath components add constructively or destructively. Fast fading changes within a short movement/time because of multipath and Doppler; slow fading follows shadowing. A channel is flat fading when signal bandwidth is well below coherence bandwidth, so all spectral components share one gain; it is frequency-selective when delay spread resolves paths and causes ISI. A useful estimate is \(B_c\approx1/(5\tau_{rms})\) Hz. Relative motion causes

At \(900\,\text{MHz}\) and \(2\,\text{km}\), \(L_{fs}\approx97.55\,\text{dB}\). A \(72\,\text{km/h}=20\,\text{m/s}\) mobile has \(\lambda=0.333\,\text{m}\) and maximum Doppler about \(60\,\text{Hz}\). Wireless links enable mobility and rapid coverage but must combat these losses using link margin, diversity/MIMO, equalization, coding, interleaving, power control and adaptive modulation; no single free-space calculation represents a cluttered fading route.

Practice target: 9 minutes; draw all propagation paths, state both loss laws, classify fading and complete one FSPL or Doppler check.