Space Information Networks

With the development of satellite networks, new generation of low earth orbit (LEO) satellite constellation network composed of mega satellites has emerged. The ground stations (GS) deployment has the key impact on the system throughput for LEO satellite constellation network. However, the GS deployment planning is challenging due to the spatial and temporal distribution of traffic demands and the time-varying topology of the LEO satellites. To solve this problem, an iterative GS deployment based on marginal revenue maximization (IGSD-MRM) is proposed. The key idea is to select the geographical location, which can achieve the maximum marginal revenue, as current GS deployment location in each iteration. The simulation results show the effectiveness of the proposed IGSD-MRM algorithm in throughput improvement under different GSs numbers and feeder link antennas.

In order to alleviate the impact of network congestion on the spatial network running traditional contact graph routing (CGR) algorithm and DTN protocol, we propose a flow intelligent control method based on deep convolutional neural network (CNN). The method includes two stages of offline learning and online prediction to intelligently predict the traffic congestion trend of the spatial network. A CGR update mechanism is also proposed to intelligently update the CGR to select a better contact path and achieve a higher congestion avoidance rate. The proposed method is evaluated in the prototype system. The experimental results show that it is superior to the existing CGR algorithm in terms of transmission delay, receiver throughput and packet loss probability.

To meet the needs of step-by-step construction of inclined geosynchronous orbit (IGSO)/medium earth orbit (MEO) constellation in the future and to improve the robustness of the constellation in case of losing connections of some satellites, this paper studies the incomplete IGSO/MEO constellation network topology design technology, analyses the inter-satellite link accessibility, and proposes a multilayer satellite network link-building strategy based on dynamic programming. This paper also designs the methods of link-building for intra-layer links among IGSO satellites, intra-layer links among MEO satellites and inter-layer links between IGSO satellites and MEO satellites. The simulation results show that this method can well construct dynamic inter-satellite network topology with less link switches, higher network transmission bandwidth and wider access user coverage.

This paper reviews the research progress of capability assessment of the networking information-centric system of systems. The concept, characteristics and capability structure of the networking information-centric system of systems are summarized, and the challenges of capability assessment are analyzed. For the problem that the nonlinear characteristics of networking information-centric system of systems capabilities are difficult to evaluate, the idea of using machine learning methods to solve this problem is proposed, and the characteristics and applications of these methods are analyzed.

Since 2013, China’s supercomputer has been ranked first in the global supercomputer rankings, and now the United States has returned to its peak, engineers at the US Department of Energy’s Oak Ridge National Laboratory released “Summit,” a supercomputer with powerful performance that surpasses the current record holder: China’s Shenwei·Taihu Light. AI has played an important role in recent development of supercomputer systems. In this paper, we discuss the opportunities and challenges of future supercomputer systems based on the thinking of AI.

Internet of Things (IoT) is rapidly gaining ground in the scenario of modern wireless communications. The satellite system, which plays a significant role in the development of IoT, is expected to cooperate with the terrestrial components to provide a complementary service. In order to better study the combination of satellite and terrestrial IoT networks, building a simulation environment based on software defined network (SDN) and network function virtualization (NFV) technology to create a real and reliable large-scale test environment is of great significance. This paper provides an architecture of IoT network in a combined low earth orbit (LEO) satellite-terrestrial structure embracing SDN technologies. A semi-physical simulation platform utilizing SDN and NFV for satellite-based IoT Network is demonstrated. Last but not least, two use cases of this platform are discussed. Simulations of satellite routing algorithms and combined LEO-terrestrial mobile network prove the validity and authenticity of the platform.

A Connected Dominating Set (CDS) is a useful method for degrading major routing and forwarding operations in the network, which is widely applied in the mobile ad hoc networks. In a flying ad hoc network (FANET), however, high dynamics of nodes produce considerably large challenges on the topology maintenance due to rapidly time-varying connections between nodes, which will lead to a huge computation latency and overheads if exploiting current CDS algorithms. In this paper, therefore, we proposed a connection estimation-based topology control mechanism to achieve efficient maintenance of connectivity in the network. In particular, the proposed algorithm could provide a stable and effective virtual backbone sub-net in a fast changing topology of FANET, by flexibly scheduling multiple Minimum Connected Dominating Sets (MCDS) with a very efficient method. The simulation results show that, compared with typical single CDS method, the proposed algorithm presents better performances in obviously dynamic environments with respect to updating counts and rate of successful updates.

Recently the topology design of multi-layer satellite network has drawn much attention from researchers due to the application of laser link in inter-satellite links (ISLs). Snapshot division is a common method of dynamic topology design, which splits the time-varying topology into a sequence of static snapshots. However, previous topology design researches based on snapshot division seldom consider multi-objective optimization and practical link parameters, such as link capacity, degree constrains, and link switch time. In this paper, a novel topology design method aiming to concurrently minimize both link switch time and network average end-to-end delay is proposed. The constraints of degree, visibility, and connectivity are considered in the proposed topology design method. Simulation results show that the proposed strategy has a better performance compared to traditional topology design methods for multi-layered optical satellite networks.

In the near future, the Space Information Network (SIN) will evolve into Space-earth Integrated Network, which will realize the interconnection among the Space-based Network, traditional Internet and Mobile Communication Network. The construction of the future SIN will provide the ability of global real-time communication and comprehensive information services, which makes it possess a wide application prospect. The future SIN will realize inter-satellite routing and its network capacity will be flexible, its network topology and link status will change dynamically as well, which will increase the complexity of the network management. In addition, the openness of communication link and the resources limitation of the devices will both bring security threats to the SIN. We study on the characteristics of future SIN and present a dynamic security protection architecture based on SDN (Software Defined Network). We study on the trusted authentication mechanism of dynamic networking entities and the technique of path optimization and risky-path isolated transmission. At last, we design a network intelligent security management mechanism at the top level. Our scheme can greatly improve the security of future SIN.

The space information networks provide a rich space, time, frequency spectrum resources, meet all kinds of scene mission requirements, especially the rapid development of information technology and the interaction of human life fusion, and the global data presents the characteristics of explosive growth and massive convergence, artificial intelligence has advantages such as flexibility, adaptability and low robustness in the direction of information fusion. On the basis of studying the framework of space-based information network, an integrate task management and control mode based on big-data driven space-based information networks and Internet of things is proposed. Artificial intelligence technology is used to solve the problem of the front-end requirements of task management and control, and the ratio of resource utilization to actual profit of joint information network load points is improved, at the same time, it lays a foundation for realizing autonomous task planning.

Discrete global grid divides the earth’s surface into approximately equal units, which is mainly used for efficient processing and visualization of the earth data. In this paper, the discrete global grid technology is introduced into the optimization of satellite-ground coverage in space-ground integrated networks. The greedy time sequence selection algorithm for satellite grid coverage is proposed to decrease the number of coverage handoff between grids and satellites, finally promotes the performance of data transmission in space-ground integrated networks.

In order to ensure the reliability of data transmission and meet the QoS (quality of service) requirements of different users, this paper analyzes the invulnerability of space-based early-warning system information network based on data transmission, taking the space-based early-warning system information network as the carrier. First of all, this paper classifies the network invulnerability. Secondly, according to the damage of the network under different circumstances, the author designs the corresponding routing strategies of invulnerability and compares it with the traditional routing algorithm. The simulation results show that the routing strategies of invulnerability proposed in this paper not only satisfies QoS requirements of different users, but also has good performance in packet loss rate, end-to-end delay and other aspects because it considers many aspects of network damage.

This paper introduces four space information network protocol researched and applied internationally: space IP protocol, CCSDS protocol, protocol combining CCSDS and TCP/IP, and delay/interrupt tolerant network (DTN) protocol. The protocol architecture, protocol components and functions are introduced and analyzed. At the same time, the method of performance improvement of space satellite communication TCP protocol is briefly introduced.

In recent years, satellite communication has becoming popular and develop very fast around the world. Recently, the fifth-generation mobile communication (5G) is about to enter the commercial use. As key characteristic of the 5G communication system, the integration of satellite communication and ground 5G has become a new hotspot. In this paper, we propose the summary of 5G in space based view, first of all, we introduce the status of 5G communication system, and its convergence to space wireless communication system. Then we analyze the development trends of satellite 5G systems, and then gives the initial idea of satellite 5G fusion, including architecture design, air interface design, based on SDN/NFV network virtualization deployment and protocol optimization, and discussed the possible problems of fusion; finally, the key technologies of satellite 5G fusion are sorted with opportunities and challenges.

With the continuous deployment of Ka multi-beam equipment, the number of satellites on that can be managed by ground system simultaneously is increasing. For the navigation constellation, the number of optical nodes is increased, which further improves the navigation accuracy and time synchronization accuracy. More time slots are available to enable the navigation constellation to provide better extended service functions. However, the sharp increase in the number of visible satellites in China leads to more complex intersatellite visibility relations, which leads to the optimization of intersatellite link planning. In particular, the calculation amount of PDOP between satellites increases exponentially, and the traditional algorithm can no longer satisfy the real-time link planning. In the paper, the k-means method is used to replace the inter-satellite distance with the angle difference. By clustering constellation satellites, the fast calculation of inter-satellite PDOP value is realized and the inter-satellite link building algorithm is optimized, which provides a solution for the rapid inter-satellite link planning after the deployment of Ka multi-beam equipment.

In view of the large number of in-orbit satellites currently in China and the shortage of ground TT&C and control resources, combining with the networking requirements of China’s navigation constellation and the current situation and development of the ground station network, this paper proposes two inter-satellite strategies based on the various service requirements of the navigation system, and carries out the utilization of ground resources separately. Through simulation analysis, the validity of the strategy is verified, and the optimal deployment schedule of ground resources is proposed, which provides effective assistant decision for the optimization of the network scheme and satellites management next.

In this paper, we focus on the satellite occurrence probability in the earth station’s visual field for mega-constellation systems. Based on the configuration of Walker constellation, we deduce the geometric relationship between the satellite phase interval and the line-of-sight angle of earth station at different elevation angles, and then analyze the law between constellation scale and line-of-sight angle of adjacent satellites. According to the given line-of-sight angle, the visual field of earth station is divided into several regions, and the satellite occurrence probability in each region can be derived. The analysis indicates that, with the expansion of the constellation scale, satellite occurrence probability in each region increases and approaches to 1. Based on the above analysis, we propose an access and handover scheme suitable for mega-constellation systems. It is assumed that the earth station is equipped with a wide shaped beam, which is pointing to a certain fixed region. There is always a visible satellite in the beam, through which the earth station connects to the constellation system and it can avoid the calculation burden of extrapolating a large number of satellite orbits. Simulation results demonstrate that the variations of space-earth link distance, radial velocity and elevation angle are extremely slight. Therefore, our proposed scheme can be well applied in the mega-constellation with frequent handovers, which can reduce the overhead of time and frequency calibration.

In this paper, we propose a distributed cooperative sensing strategy for multi-satellites based on coalition formation games. To reduce the complexity, we introduce adjunct utility function to achieve steady-state coalition. The simulation results show that the merger-division-adjust (MDA) algorithm proposed in this paper can reduce the probability of missed detection compared with the non-cooperative algorithm. Compared with merger and division (MD) algorithm, MDA algorithm can reduce the probability of false alarm, indicating that the proposed strategy has better spectrum sensing performance and more suitable for satellites.

The wireless transmissions of the satellite communications system are vulnerable to the interferences. In realistic systems, the satellite communications system faces three kinds of interference scenarios. Built upon the analysis on the possibly occurred three interference scenarios, we propose interference avoidance oriented cognitive satellite communications architecture such that the reliable communications can be guaranteed under the spectrum resource contentions and complex electromagnetic environments. Following the developed architecture, we further analyze the corresponding interference avoidance approaches. The work conducted in this paper could support the construction and development of future cognitive satellite communications systems.

The conventional rate compatible modulation (RCM) uses the high-order quadrature amplitude modulation (QAM) signal for rateless transmitting. The large size complex constellation with near-Gaussian probability mass function (PMF) produces high peak-to-average power ratio (PAPR). This paper presents a new method aimed at solving the PAPR problem associated with RCM. We transform the RCM signal to a constant envelope signal through concatenating RCM with continuous phase modulation (CPM), which decreases the PAPR to 0 dB. At the receiver, the serial iterative demodulating and decoding procedure is designed to improve the performance of the system. In the presence of nonlinear power amplifier, we simulate the bit error rate and spectral efficiency of the RCM-CPM and the RCM-QAM in additive Gaussian noise channels. The simulation results demonstrate that RCM-CPM outperforms RCM-QAM with input back-off, especially the performance advantage is about 3 dB at high signal-to-noise ratios (SNRs).

A hybrid precoding scheme is proposed for high altitude platform (HAP) massive multiple-input multiple-output (MIMO) systems to obtain the radio frequency (RF) precoder and the baseband precoder with limited RF chains. We first exploit duality theory to derive the relation between the statistical channel state information (SCSI) and RF precoder, which is selected from a predefined codebook. Then, the baseband precoder is attained by zero forcing (ZF) based on the instantaneous effective channel matrix. A distinct performance gain is achieved by the proposed scheme according to the numerical results.

In order to improve the security and concealment of satellite communication, an anti-interception technology based on weighted fractional Fourier transform (WFRFT) and transform domain communication system (TDCS) is proposed. WFRFT replaces FFT/IFFT and CCSK frequency-domain mapping replaces time-domain modulation in TDCS. Moreover, the receiving basis function is designed according to the principle of equal gain combining. The simulation results show that, WFRFT-TDCS anti-eavesdropping technology can be applied to satellite transceiver of spectrum mismatch, as well as maintain system complexity and enhance anti-scanning performance. Even if the SNR is greatly increased, unauthorized receiver cannot demodulate the signal correctly. For authorized receiver, SNR loss is less than 1 dB when the transmission spectrum availability is more than 80%.

In view of the increasing on-board processing capacity, this paper investigates the possibility of the communication reconnaissance on LEO constellation satellite platforms, and proposes a wireless signal recognition algorithm based on deep learning. The proposed algorithm visualizes the wireless signal as a picture based on the basic digital signal processing, as a result, the signal recognition problem is subtly transferred to an object detection problem recurring in the field of Computer Vision (CV). Then, it co-opts deep learning models in CV field in order to realize the end-to-end signal recognition and improve the performance. Validating results on the field-collected signal dataset with 12 types and 4740 samples show that, the algorithm can effectively identify the waveform types and time/frequency coordinates of communication signals with the precision 89%, which is 40% higher than traditional algorithms.