Resource Allocation for 5G-UAV Based Emergency Wireless Communications

I. INTRODUCTION

• UAV-based emergency networks face unclear channel models and limited resources, while disaster-driven terrain changes make channel characteristics difficult to represent.The introduction identifies changing shadowing and the distinct conditions of command-center–UAV and UAV–user links as modeling challenges.Source: §I · pp. 1–2
• The paper models multiple IRS-equipped 5G-UAVs relaying command-center signals to trapped users over heterogeneous emergency channels.The proposed Fisher-Snedecor F composite model accounts for reflectors, path-loss exponent, fading, and shadowing parameters.Source: §I · p. 2
• The authors propose channel-adaptive power allocation and joint bandwidth-power optimization to improve capacity and energy efficiency.They compare capacities for heterogeneous-F-adapted and general-fading-adapted allocation, and assess energy efficiency for IRS-equipped versus relay-based 5G-UAVs.Source: §I · p. 2
• Numerical results report higher capacity and energy efficiency for heterogeneous-F-adapted resource allocation than for traditional channel-model-adapted schemes.The stated implication is better communications service for post-disaster areas.Source: §I · p. 2

II. THE 5G-UAV BASED EWC SYSTEM MODEL

• The modeled network has one on-site command center, K IRS-equipped 5G-UAVs with N passive reflectors each, and irregularly distributed trapped users.Each user receives an individually assigned frequency band or time slot; UAVs hover between the command center and user to reflect signals using phase shifts.Source: §II · p. 2
• The user’s received signal is the superposition of K UAV-path signals, with the available bandwidth divided into K UAV sub-channels.The model includes the transmitted signal, IRS-element phase shifts, and additive white Gaussian noise.Source: §II · pp. 2–3
• The command-center–UAV links are modeled mainly with path loss, while UAV–user links use Fisher-Snedecor F fading to capture fading and shadowing.The distinction reflects reduced shadowing on the positioned command-center–UAV links and changing propagation conditions around trapped users.Source: §II · p. 3
• The model assigns each sub-channel an independent Fisher-Snedecor F random variable with fading parameter m, shadowing parameter ms, and mean power.The passage identifies the probability density function and defines the beta and gamma functions used in its expression.Source: §II · p. 3
• The heterogeneous channel gain h aggregates UAV-path gains, while the sub-channel SNR depends on path loss, reflector gain, noise variance, bandwidth, and transmit power.The reflector contribution is the sum power gain of N elements; the stated noise variance is N0B.Source: §II · p. 4

III. RESOURCE ALLOCATION SCHEMES FOR MULTIPLE 5G-UAV BASED EWC NETWORKS

• The section derives a heterogeneous F composite fading model and examines how channel parameters and the number of UAVs affect its power-gain probability density.It then studies the model-adapted power allocation scheme for maximum ergodic capacity and how allocated power varies with gain, fading, and shadowing parameters.Source: §III · p. 4
• The section develops a channel-adaptive optimal power allocation scheme to obtain maximum ergodic capacity for 5G-UAV emergency networks.The analysis varies the power-gain, fading, and shadowing parameters.Source: §III · p. 4
• A joint bandwidth-power optimization algorithm is developed to maximize ergodic capacity, and its convergence is analyzed.Source: §III · p. 4
• The section analyzes energy efficiency for IRS-based 5G-UAVs and relay-based 5G-UAVs.Source: §III · p. 4

A. The Heterogeneous F Composite Fading Channel

• The model defines PDFs for the k-th sub-channel power gain hk and the overall heterogeneous F composite fading channel power gain h.The analysis then examines how channel parameters and the number of reflectors affect the overall PDF.Source: A. The Heterogeneous F Composite… · p. 4
• More 5G-UAVs K shift the power-gain PDF toward larger gains, increasing average power gain h.Fig. 2 presents the PDF against h and K.Source: Figure 2 · p. 4, A. The Heterogeneous F Composite… · pp. 4–5
• More reflectors N shift the PDF toward larger power gains, and the reported power gain increases as N increases.Fig. 3 presents the PDF against h and N; Fig. 4 compares the PDF over h for different reflector counts.Source: Figure 3 · p. 4, A. The Heterogeneous F Composite… · p. 5, Figure 4 · p. 5
• Average power gain increases as path-loss exponent α decreases and shadowing parameter ms increases.As ms increases, the range of achievable power gain becomes more concentrated.Source: A. The Heterogeneous F Composite… · p. 5, Figure 6 · p. 5, Figure 5 · p. 5

B. The Heterogeneous F Composite Fading Channel Adapted Optimal Power Allocation Scheme

• The derivation treats P1 as a strictly convex optimization problem, using concavity of its objective and linear power constraints.The solution process forms a Lagrangian and applies KKT conditions.Source: B. The Heterogeneous F Composite… · p. 6
• The optimal power allocation for P1 is derived in closed form, with zero power when h is below outage threshold h0.The derivation uses the Lagrangian and KKT conditions; the paper also derives an asymptotic expression for heavy-shadowing cases with ms ≤ m.Source: B. The Heterogeneous F Composite… · pp. 6–7
• The power-allocation scheme yields the maximum channel capacity, expressed as Eq. (20).The section identifies [ξ]+ as max{ξ, 0}.Source: B. The Heterogeneous F Composite… · p. 6
• Allocated power rises with power gain and fading parameter m, and rises as shadowing parameter ms decreases.The paper associates lower ms with heavier shadowing, requiring more power to satisfy basic emergency-communication requirements.Source: B. The Heterogeneous F Composite… · p. 7

C. Joint Bandwidth-Power Allocation Scheme

• The joint bandwidth-power problem is formulated as a strictly convex optimization problem.The objective is described as concave in the allocated bandwidth and power, while the resource terms are linear.Source: C. Joint Bandwidth-Power Allocation… · p. 8
• The scheme allocates bandwidth and power across K UAV subchannels according to instantaneous channel state to maximize ergodic capacity.The motivation is that disaster-driven channel states change dynamically and received SNR can remain relatively low even with cooperating UAVs.Source: C. Joint Bandwidth-Power Allocation… · pp. 8–9
• The algorithm combines KKT-based sub-optimal solutions under given bandwidth or power limits with gradient descent to find the optimal solution.Iterations stop when capacity increments fall below a threshold or the maximum iteration count is reached.Source: C. Joint Bandwidth-Power Allocation… · p. 9

D. Energy Efficiency Analysis

• Relay power consumption includes circuit blocks, the relay FPGA, and the power amplifier, while IRS consumption includes the IRS FPGA and PIN diodes.The relay power model also includes power-conversion efficiency.Source: D. Energy Efficiency Analysis · p. 9
• The comparison defines relay and IRS energy efficiencies using their respective ergodic capacities and power consumptions.The IRS capacity is considered under the same K-subchannel case as the relay network.Source: D. Energy Efficiency Analysis · p. 10
• IRS-assisted 5G-UAVs have lower total power consumption because PIN diodes replace RF chains and consume little power.The paper identifies this low consumption as the main reason for IRS energy-efficiency superiority over traditional relays in resource-limited emergency communications.Source: D. Energy Efficiency Analysis · p. 10

IV. PERFORMANCE EVALUATIONS

• The heterogeneous F composite fading channel adapted scheme outperforms traditional channel adapted power allocation in capacity comparisons with fading and shadowing parameters.The fading comparison fixes ms=1; the shadowing comparison fixes m=5.Source: §IV · p. 11, Figure 11 · p. 12, Figure 12 · p. 12
• With close or equal sub-channel gains, allocated bandwidth and power converge to fixed values; equal-gain channels receive about 65MHz bandwidth each.For close gains, the example allocation for 5G-UAV 1 converges around iteration 4500 to 60MHz and 9mW.Source: Figure 13 · p. 12, §IV · p. 12, Figure 14 · p. 12
• With diverse channel gains, the best sub-channel receives the most resources, and the allocation and capacity converge to fixed values.For h1=4, h2=5, and h3=6, 5G-UAV 1's bandwidth and power converge to zero around iteration 4500.Source: Figure 15 · p. 13, §IV · p. 13
• The joint bandwidth-power allocation scheme achieves higher capacity than the fixed-bandwidth scheme by adapting bandwidth to instantaneous channel state.The comparison is shown in Fig. 16, with adaptive resource allocation validated by Figs. 13–15.Source: Figure 16 · p. 13, §IV · p. 13
• Under the given power constraint, IRS-assisted 5G-UAVs have higher energy efficiency than 5G-UAVs without IRSs.The comparison evaluates N=8 and N=16 reflectors corresponding to active PIN diodes.Source: Figure 17 · p. 13, §IV · p. 13
• IRS-based 5G-UAVs have higher energy efficiency than relay-based 5G-UAVs and MIMO UAVs with the same antenna count.The paper attributes the MIMO comparison to lower PIN-diode power consumption than power-amplifier consumption.Source: §IV · pp. 13–14, Figure 18 · p. 14, Figure 19 · p. 14

V. CONCLUSIONS

• The proposed heterogeneous F composite fading model characterizes path loss, reflector count, fading, and shadowing in complex EWC scenarios.The paper develops optimal power allocation and joint bandwidth-power allocation schemes based on this model.Source: §V · p. 14
• Numerical results show higher capacity for the proposed model-adapted power allocation than for traditional fading-model-adapted schemes.Source: §V · p. 14
• The joint bandwidth-power allocation scheme converges and outperforms fixed-bandwidth allocation; IRS-based 5G-UAVs also exceed relay-based 5G-UAVs in energy efficiency.The paper concludes that IRS-based UAVs are conducive to restoring communications in post-disaster areas.Source: §V · p. 14

APPENDIX A PROOF OF LEMMA 1

• The proof begins with the PDF of gk derived according to Eq. (2).Source: Appendix A · p. 14
• The proof rewrites the density using Meijer’s G-function, with the hypergeometric function identified as 2F1.Source: Appendix A · p. 14
• It integrates the gain density to form a CDF, then uses gk=Lαk hk to express the CDF and PDF in terms of hk.Source: Appendix A · p. 14

APPENDIX B PROOF OF LEMMA 2

• The proof obtains the moment generating function of hk and uses it to rewrite the MGF of h.Source: Appendix B · pp. 14–15
• It derives the PDF of h using the inverse Laplace transform of the MGF.Source: Appendix B · p. 15
• The resulting density is expressed with a Lauricella multivariate hypergeometric function, which reduces to a simpler expression for i.i.d. Fisher-Snedecor random variables.Source: Appendix B · p. 15

• The fragments show the sequence ds1ds2 . . . dsK.Source: Nmk) · p. 16
• One fragment places B before a parenthesized list of Nm and Nms terms.Source: Nmk) · p. 16
• Another fragment lists Nm1 through NmK, separated by commas.Source: Nmk) · p. 16