In recent years, much work has been done on processing of wireless spectrum data involving machin... more In recent years, much work has been done on processing of wireless spectrum data involving machine learning techniques in domain-related problems for cognitive radio networks, such as anomaly detection, modulation classification, technology classification and device fingerprinting. Most of the solutions are based on labeled data, created in a controlled manner and processed with supervised learning approaches. However, spectrum data measured in real-world environment is highly nondeterministic, making its labeling a laborious and expensive process, requiring domain expertise, thus being one of the main drawbacks of using supervised learning approaches in this domain. In this paper, we investigate the use of self-supervised learning (SSL) for exploring spectrum activities in a real-world unlabeled data. In particular, we compare the performance of two SSL models, one based on a reference DeepCluster architecture and one adapted for spectrum activity identification and clustering, and a baseline model based on K-means clustering algorithm. We show that SSL models achieve superior performance regarding the quality of extracted features and clustering performance. With SSL models we achieve reduction of the feature vectors size by two orders of magnitude, while improving the performance by a factor of 2 to 2.5 across the evaluation metrics, supported by visual assessment. Additionally we show that adaptation of the reference SSL architecture to the domain data provides reduction of model complexity by one order of magnitude, while preserving or even improving the clustering performance.
The spread of wideband services to the office and home through fixed and terrestrial mobile netwo... more The spread of wideband services to the office and home through fixed and terrestrial mobile networks will create a market for the extension of such services for people on the move and for those who work outside the office environment. Various wideband terrestrial cellular systems will serve users at both local and global level, although the scope for deployment of such systems in the early phase (2000+) will be limited to densely populated and affluent areas, resulting in a potential requirement for wireless access technologies which can connect users to the network in other areas.
To enable throughput rates in the terabit-persecond range, future satellite communications will h... more To enable throughput rates in the terabit-persecond range, future satellite communications will have to utilize a large available bandwidth in the Q-band and above. At these high frequencies, the signal is severely degraded and attenuated by tropospheric phenomena, especially rainfall, which reduces the availability of satellite communication link and degrades the quality of service. Site diversity is an efficient technique used to mitigate the rain attenuation. In this paper, we evaluate the performance of the large-scale site diversity system consisting of ground stations in Ljubljana and Budapest based on long-term satellite signal measurement campaigns of Alphasat at 39.402 GHz. The measured attenuation time series are statistically analyzed and presented for a period of one year and compared with three representative prediction models.
2008 IEEE International Workshop on Satellite and Space Communications, 2008
For satellite systems offering broadband and telecom services at Ka band frequencies and above Fa... more For satellite systems offering broadband and telecom services at Ka band frequencies and above Fade Mitigation Techniques (FMT), in particular Adaptive Modulation and Coding (ACM), are essential to improve the system capacity. One of the important aspects of ACM from a physical and MAC layer perspective is the choice of transmission modes to be used during the system operation that would achieve the optimal trade-off between the system complexity and performance. This paper deals with one such methodology of selecting the appropriate set of transmission modes (MODCOD subset) from a physical layer point of view for both DVB-S2 and extended DVB-RCS systems. Optimizing the achievable spectral efficiency with respect to implementation complexity has been the main focus of the proposed methodology. In particular, we show that subsets with 8 and 5 transmission modes provide best trade-off between the complexity of the ACM based system and the achieved average spectral efficiency for the forward and return link of the DVB-S2/RCS system respectively.
IEEE Transactions on Aerospace and Electronic Systems, 2007
Adaptive per-hop routing in the intersatellite link (ISL) network of a packet-switched, nongeosta... more Adaptive per-hop routing in the intersatellite link (ISL) network of a packet-switched, nongeostationary satellite system is addressed. In particular, a traffic class dependent (TCD) routing is proposed with different optimization criteria for different traffic classes. Furthermore, for delay sensitive traffic, performance is enhanced by an exponential smoothing link-cost function, which reduces traffic load oscillations in the network. The performance of the proposed procedures is evaluated in two different traffic scenarios using an appropriate simulation model.
Orthogonal frequency division multiplexing (OFDM) system with frequency domain equalization (FDE)... more Orthogonal frequency division multiplexing (OFDM) system with frequency domain equalization (FDE) requires reliable channel estimation (CE). OFDM has a problem of high peak-to-average power ratio (PAPR), which makes it very sensitive to nonlinear distortions, affecting the channel estimation accuracy. In this paper, we investigate the effect of the nonlinearity to the the OFDM system with pilot-assisted CE based on time or frequency division multiplexed (TDM/FDM) pilot. A closed-form bit error rate (BER) expressions for OFDM system are derived in a nonlinear and frequency-selective fading channel. The analysis is based on the Gaussian approximation of the nonlinear noise, which is also confirmed by computer simulation. Our results in terms of BER and mean square error (MSE) show, that FDM-pilot based CE is more sensitive to nonlinear distortions as compared to CE based on TDM-pilot.
This paper is concerned with the development of an aggregate traffc intensity model suitable for ... more This paper is concerned with the development of an aggregate traffc intensity model suitable for the performance evaluation of the Intersatellite Link (ISL) network in global satellite systems, particularly applicable to support the dimensioning of satellite networks and for computer simulations of various procedures in the ISL network segment. The model consists of four modules considering global distribution of traffic sources and destinations, temporal variations of traffic sources' intensity, traffc flow patterns between regions and statistical behavior of aggregated traffic sources. To support universal utilization of the proposed aggregate traffic intensity model, the individual modules need to he efficiently integrated and implemented as a highly parameterized software tool.
We propose a framework for dynamic composition of communication services which is well suited for... more We propose a framework for dynamic composition of communication services which is well suited for facilitating research and prototyping on real experimental infrastructures of remotely configurable embedded devices. By using the concept of composeability, our framework supports modular component development for various networking functions, therefore promoting code re-use. The framework consists of four components: the physical testbed, the module library, the declarative language and the workbench. Its reference implementation, ProtoStack, developed using semantic web technologies, supports remote experimentation on sensor platform based infrastructure, thus being well suited also for experimenters that do not poses their own physical experimentation infrastructure. We illustrate how ProtoStack supports research in service oriented networks and a cognitive networking. The cost of increased flexibility and prototyping speed of the protocol stack is paid in terms of increased memory footprint, processing speed and energy consumption. Compared to the most related non composable approach, the CRime library used by ProtoStack has 16 to 17% larger footprint, it takes 2.4 times longer to execute an open¨send¨recv¨close sequence and consumes 1.6% more power in doing so. Even though with ProtoStack more resources are consumed by the node, the tradeoff in terms of prototyping speed pays it off.
Proceedings of the 1st ACM workshop on Networking, systems, and applications for mobile handhelds - MobiHeld '09, 2009
Service oriented access in a multi-application, multi-access network environment poses interestin... more Service oriented access in a multi-application, multi-access network environment poses interesting research challenges. One of these challenges refers to cross-layer interoperability among technologies. In this poster, we introduce a knowledge base (KB) which contains local (user terminal specific) knowledge that enables pro-active network selection by translating technology specific parameters to higher-level, more abstract parameters. We implemented a prototype which makes use of semantic technology (namely ResearchCyc) for creating the elements of the KB: the ontology, the concepts, facts and rules. The system implements technologyspecific QoS parameters mapping according to the IEEE 802.21 draft standard recommendation.
2006 International Workshop on Satellite and Space Communications, 2006
Abstract High altitude platform (HAP) system offers establishment of communication links with pre... more Abstract High altitude platform (HAP) system offers establishment of communication links with predominantly line-of-sight (LOS) conditions due to high elevation angles. However, in mobile scenario, the transmitted signal is still subject to blocking due to obstacles between ...
2014 11th International Symposium on Wireless Communications Systems (ISWCS), 2014
A relatively small number of testbeds based on constrained devices use a wireless management netw... more A relatively small number of testbeds based on constrained devices use a wireless management network, mostly because of the unreliable communication it enables. However, in some cases, such management networks are the only option due to the target location of such testbeds: outdoors, on light posts, buildings, etc. In this paper, we analyze the design trade-offs encountered when designing a wireless management network for testbeds based on constrained devices. First, we identify two use cases and the functionality needed by the management network in supporting them. Next, we discuss ways of providing the desired functionality and illustrate the decisions we took for designing and implementing the management network for the extension of the LOG-a-TEC testbed together with an initial evaluation. The analysis and the adopted decisions resulted in a management network that is separated from the experimental network providing improved application throughput, together with smaller application level updates/reconfiguration size that significantly shorten the time required to set up a new experiment.
In recent years, much work has been done on processing of wireless spectrum data involving machin... more In recent years, much work has been done on processing of wireless spectrum data involving machine learning techniques in domain-related problems for cognitive radio networks, such as anomaly detection, modulation classification, technology classification and device fingerprinting. Most of the solutions are based on labeled data, created in a controlled manner and processed with supervised learning approaches. However, spectrum data measured in real-world environment is highly nondeterministic, making its labeling a laborious and expensive process, requiring domain expertise, thus being one of the main drawbacks of using supervised learning approaches in this domain. In this paper, we investigate the use of self-supervised learning (SSL) for exploring spectrum activities in a real-world unlabeled data. In particular, we compare the performance of two SSL models, one based on a reference DeepCluster architecture and one adapted for spectrum activity identification and clustering, and a baseline model based on K-means clustering algorithm. We show that SSL models achieve superior performance regarding the quality of extracted features and clustering performance. With SSL models we achieve reduction of the feature vectors size by two orders of magnitude, while improving the performance by a factor of 2 to 2.5 across the evaluation metrics, supported by visual assessment. Additionally we show that adaptation of the reference SSL architecture to the domain data provides reduction of model complexity by one order of magnitude, while preserving or even improving the clustering performance.
The spread of wideband services to the office and home through fixed and terrestrial mobile netwo... more The spread of wideband services to the office and home through fixed and terrestrial mobile networks will create a market for the extension of such services for people on the move and for those who work outside the office environment. Various wideband terrestrial cellular systems will serve users at both local and global level, although the scope for deployment of such systems in the early phase (2000+) will be limited to densely populated and affluent areas, resulting in a potential requirement for wireless access technologies which can connect users to the network in other areas.
To enable throughput rates in the terabit-persecond range, future satellite communications will h... more To enable throughput rates in the terabit-persecond range, future satellite communications will have to utilize a large available bandwidth in the Q-band and above. At these high frequencies, the signal is severely degraded and attenuated by tropospheric phenomena, especially rainfall, which reduces the availability of satellite communication link and degrades the quality of service. Site diversity is an efficient technique used to mitigate the rain attenuation. In this paper, we evaluate the performance of the large-scale site diversity system consisting of ground stations in Ljubljana and Budapest based on long-term satellite signal measurement campaigns of Alphasat at 39.402 GHz. The measured attenuation time series are statistically analyzed and presented for a period of one year and compared with three representative prediction models.
2008 IEEE International Workshop on Satellite and Space Communications, 2008
For satellite systems offering broadband and telecom services at Ka band frequencies and above Fa... more For satellite systems offering broadband and telecom services at Ka band frequencies and above Fade Mitigation Techniques (FMT), in particular Adaptive Modulation and Coding (ACM), are essential to improve the system capacity. One of the important aspects of ACM from a physical and MAC layer perspective is the choice of transmission modes to be used during the system operation that would achieve the optimal trade-off between the system complexity and performance. This paper deals with one such methodology of selecting the appropriate set of transmission modes (MODCOD subset) from a physical layer point of view for both DVB-S2 and extended DVB-RCS systems. Optimizing the achievable spectral efficiency with respect to implementation complexity has been the main focus of the proposed methodology. In particular, we show that subsets with 8 and 5 transmission modes provide best trade-off between the complexity of the ACM based system and the achieved average spectral efficiency for the forward and return link of the DVB-S2/RCS system respectively.
IEEE Transactions on Aerospace and Electronic Systems, 2007
Adaptive per-hop routing in the intersatellite link (ISL) network of a packet-switched, nongeosta... more Adaptive per-hop routing in the intersatellite link (ISL) network of a packet-switched, nongeostationary satellite system is addressed. In particular, a traffic class dependent (TCD) routing is proposed with different optimization criteria for different traffic classes. Furthermore, for delay sensitive traffic, performance is enhanced by an exponential smoothing link-cost function, which reduces traffic load oscillations in the network. The performance of the proposed procedures is evaluated in two different traffic scenarios using an appropriate simulation model.
Orthogonal frequency division multiplexing (OFDM) system with frequency domain equalization (FDE)... more Orthogonal frequency division multiplexing (OFDM) system with frequency domain equalization (FDE) requires reliable channel estimation (CE). OFDM has a problem of high peak-to-average power ratio (PAPR), which makes it very sensitive to nonlinear distortions, affecting the channel estimation accuracy. In this paper, we investigate the effect of the nonlinearity to the the OFDM system with pilot-assisted CE based on time or frequency division multiplexed (TDM/FDM) pilot. A closed-form bit error rate (BER) expressions for OFDM system are derived in a nonlinear and frequency-selective fading channel. The analysis is based on the Gaussian approximation of the nonlinear noise, which is also confirmed by computer simulation. Our results in terms of BER and mean square error (MSE) show, that FDM-pilot based CE is more sensitive to nonlinear distortions as compared to CE based on TDM-pilot.
This paper is concerned with the development of an aggregate traffc intensity model suitable for ... more This paper is concerned with the development of an aggregate traffc intensity model suitable for the performance evaluation of the Intersatellite Link (ISL) network in global satellite systems, particularly applicable to support the dimensioning of satellite networks and for computer simulations of various procedures in the ISL network segment. The model consists of four modules considering global distribution of traffic sources and destinations, temporal variations of traffic sources' intensity, traffc flow patterns between regions and statistical behavior of aggregated traffic sources. To support universal utilization of the proposed aggregate traffic intensity model, the individual modules need to he efficiently integrated and implemented as a highly parameterized software tool.
We propose a framework for dynamic composition of communication services which is well suited for... more We propose a framework for dynamic composition of communication services which is well suited for facilitating research and prototyping on real experimental infrastructures of remotely configurable embedded devices. By using the concept of composeability, our framework supports modular component development for various networking functions, therefore promoting code re-use. The framework consists of four components: the physical testbed, the module library, the declarative language and the workbench. Its reference implementation, ProtoStack, developed using semantic web technologies, supports remote experimentation on sensor platform based infrastructure, thus being well suited also for experimenters that do not poses their own physical experimentation infrastructure. We illustrate how ProtoStack supports research in service oriented networks and a cognitive networking. The cost of increased flexibility and prototyping speed of the protocol stack is paid in terms of increased memory footprint, processing speed and energy consumption. Compared to the most related non composable approach, the CRime library used by ProtoStack has 16 to 17% larger footprint, it takes 2.4 times longer to execute an open¨send¨recv¨close sequence and consumes 1.6% more power in doing so. Even though with ProtoStack more resources are consumed by the node, the tradeoff in terms of prototyping speed pays it off.
Proceedings of the 1st ACM workshop on Networking, systems, and applications for mobile handhelds - MobiHeld '09, 2009
Service oriented access in a multi-application, multi-access network environment poses interestin... more Service oriented access in a multi-application, multi-access network environment poses interesting research challenges. One of these challenges refers to cross-layer interoperability among technologies. In this poster, we introduce a knowledge base (KB) which contains local (user terminal specific) knowledge that enables pro-active network selection by translating technology specific parameters to higher-level, more abstract parameters. We implemented a prototype which makes use of semantic technology (namely ResearchCyc) for creating the elements of the KB: the ontology, the concepts, facts and rules. The system implements technologyspecific QoS parameters mapping according to the IEEE 802.21 draft standard recommendation.
2006 International Workshop on Satellite and Space Communications, 2006
Abstract High altitude platform (HAP) system offers establishment of communication links with pre... more Abstract High altitude platform (HAP) system offers establishment of communication links with predominantly line-of-sight (LOS) conditions due to high elevation angles. However, in mobile scenario, the transmitted signal is still subject to blocking due to obstacles between ...
2014 11th International Symposium on Wireless Communications Systems (ISWCS), 2014
A relatively small number of testbeds based on constrained devices use a wireless management netw... more A relatively small number of testbeds based on constrained devices use a wireless management network, mostly because of the unreliable communication it enables. However, in some cases, such management networks are the only option due to the target location of such testbeds: outdoors, on light posts, buildings, etc. In this paper, we analyze the design trade-offs encountered when designing a wireless management network for testbeds based on constrained devices. First, we identify two use cases and the functionality needed by the management network in supporting them. Next, we discuss ways of providing the desired functionality and illustrate the decisions we took for designing and implementing the management network for the extension of the LOG-a-TEC testbed together with an initial evaluation. The analysis and the adopted decisions resulted in a management network that is separated from the experimental network providing improved application throughput, together with smaller application level updates/reconfiguration size that significantly shorten the time required to set up a new experiment.
Uploads
Papers by M. Mohorcic