The image task scheduling demand exhibits a thriving trend since China leverages heterogeneous low earth orbit (LEO) satellites to empower the earth observation field for Belt and Road Initiative (BRI) countries and regions. However, the high-dynamic mobility and intermittent inter-satellite links (ISLs) exacerbate communication, caching, and computing (3C) resources scarcity, while inefficient scheduling significantly deteriorates the successful transmission ratio (STR). To tackle this challenge, this paper optimizes the task scheduling strategy. To be specific, we first propose a dynamic resource mapping model (DRMM) that captures the dynamic characteristics of 3C resources through discontinuous ISLs and quantifies resources over continuous time. Based on the DRMM, we formulate an STR maximization problem under heterogeneous resource capacity constraints. A computing resources priority allocation algorithm (CRPAA) is designed to preferentially allocate onboard computing resources for task processing, thereby freeing up transmission and caching resources for additional tasks. Furthermore, the CRPAA leverages an incrementally searching slots range to reduce computational complexity and optimize the scheduling strategy, planning each task individually and clearing it promptly to prevent redundant scheduling while enhancing execution efficiency. The extensive simulation results validate that the proposed algorithm outperforms benchmark approaches.

Integrating aerial base stations (ABSs) with terrestrial base stations (TBSs) represents a promising architecture for future networks. However, challenges arise from the ABS mobility and complex interference, leading to degradation in the coverage performance, including both the average coverage quality and coverage stability. To address these challenges, we investigate the average coverage quality and coverage stability via the first- and second-order statistical properties of network spatial throughput, respectively. Our findings reveal that the inappropriate ABS deployment, especially the antenna beamwidth, causes the average coverage quality deterioration due to the co- and cross-layer interference surge induced by the overlapping coverage between ABSs and TBSs. Additionally, coverage stability experiences degradation due to the ABS mobility, particularly exacerbated by factors such as the large deployment density and circling radius as well as the small flight height and antenna beamwidth of ABSs. Moreover, it is demonstrated that there exists an optimal ABS antenna beamwidth maximizing the coverage performance. On this account, we propose an antenna beamwidth optimization algorithm as well as a time-efficient but low performance loss alternative antenna beamwidth optimization strategy, aimed at mitigating the overlapping coverage-induced interference and reducing the ABS mobility-induced impact, both of which are validated through numerical results.

Aerial base stations (ABSs) flying in the air inject wireless networks more flexibility and agility beyond ground base stations (GBSs) to respond to spatio-temporal coverage demand. To fully unlock the potential of ABSs, a high-capacity, flexible and dynamic wireless backhaul provision is necessitated and integrated access and backhaul (IAB) architecture comes into the picture. In this paper, we investigate the availability of IAB architecture in dynamic aerial-terrestrial networks in terms of coverage probability (CP) and further explore the feasible region of IAB to promote aerial-terrestrial coverage enhancement. Specifically, the results show that the capability of IAB to promote aerial-terrestrial coverage enhancement would be diminished with the increases of ABSs flight speed and GBS density. The reason is found that relying on fixed GBSs to provide dynamic backhaul for flying ABSs would come with frequent handovers, which degrades network CP. On this account, to make IAB adapt to dynamic aerial-terrestrial networks, a mobility-adaptable IAB scheme is proposed where a distance thresholdLpis set to alleviate the negative effect caused by handovers. WithLpoptimized, the coverage performance of dynamic aerial-terrestrial IAB network is shown to be increased, especially in dense GBS regime.

Exploiting the time-varying feature of wireless channels, retransmission could enable reliable data transmission. However, the growing deployment of small cell base stations (BSs) would induce significant temporal interference correlation. In consequence, once the current transmission fails, the subsequent ones are likely to fail as well. In this light, we investigate the retransmission performance in dense networks with temporally correlated interference in terms of network spatial throughput (ST). Our results reveal that the impact of temporal interference correlation on the retransmission performance critically depends on the BS density. Specifically, temporal interference correlation would cause a greater network ST attenuation when the BS density is closer to the critical density, under which network ST is maximized. Moreover, the impact of temporal interference correlation is shown to be cumulative as the number of retransmission attempts increases. Furthermore, towards efficient retransmission, we adopt and optimize a P-Activation strategy (PAS). It is shown that the optimized PAS is able to effectively mitigate the overwhelming strength and temporal correlation of interference. As a result, the retransmission performance in improving network ST is significantly enhanced in dense networks, while the variation of network ST with BS density exhibits sigmoid trend instead of the previous near-bell shape.

With the growing deployment of small cell base stations (BSs), the impact of backhaul congestion on the performance of small cell networks becomes dominant. To capture this effect, we present a conjoint framework integrating backhaul architecture with the access network. In particular, we evaluate the performance of ultra-dense networks (UDNs) in terms of network spatial throughput (ST), supposing that backhaul is conveyed to BSs through millimeter wave (mmWave) links by gateways. Notably, in contrast to diminishing to zero in previous work, network ST is shown to converge into a saturation under the given gateway density since the limited backhaul capability of gateways would stabilize the interference distribution of access network. Moreover, despite the benefit of enhancing backhaul capacity, over-deployed gateways would result in the potential intercell interference, which degrades network ST of UDN. On this account, we further study the optimization of gateway density to balance the tradeoff between enhancing backhaul capacity and mitigating intercell interference. It indicates that the application of mmWave beamforming at gateways leads to an increase in optimal gateway density, under which network ST could be further improved.

Interference serves as the most dominant factor that quantitatively and qualitatively impacts the performance of ultra-dense networks (UDN). Especially, since interference of different time slots basically comes from the same set of interfering base stations (BSs), the temporal interference correlation becomes more significant with the growing deployment of network infrastructures. In this light, we develop an analytical framework to investigate the impact of temporal interference correlation on UDN in terms of network spatial throughput (ST) in this work. In contrast to the available research, which indicates that the interference correlation is independent of BS density, we show that a growing BS density would exacerbate the influence of interference correlation on the performance of UDN. In particular, when the BS density is closer to the critical density, under which network ST could be maximized, the ST attenuation caused by temporally correlated interference is more significant. Moreover, the effect of temporal interference correlation on network ST is additive over time slots. For instance, a greater number of transmissions of hybrid automatic repeat request (HARQ) would result in a more significant effect of temporally correlated interference on network ST.