Learning manipulation skills from open surgery provides more flexible access to the organ targets in the abdomen cavity and this could make the surgical robot working in a highly intelligent and friendly manner. Teaching by demonstration (TbD) is capable of transferring the manipulation skills from human to humanoid robots by employing active learning of multiple demonstrated tasks. This work aims to transfer motion skills from multiple human demonstrations in open surgery to robot manipulators in robot-assisted minimally invasive surgery (RA-MIS) by using TbD. However, the kinematic constraint should be respected during the performing of the learned skills by using a robot for minimally invasive surgery. In this article, we propose a novel methodology by integrating the cognitive learning techniques and the developed control techniques, allowing the robot to be highly intelligent to learn senior surgeons' skills and to perform the learned surgical operations in semiautonomous surgery in the future. Finally, experiments are performed to verify the efficiency of the proposed strategy, and the results demonstrate the ability of the system to transfer human manipulation skills to a robot in RA-MIS and also shows that the remote center of motion (RCM) constraint can be guaranteed simultaneously. Note to Practitioners-This article is inspired by limited access to the manipulation of laparoscopic surgery under a kinematic constraint at the point of incision. Current commercial surgical robots are mostly operated by teleoperation, which is representing less autonomy on surgery. Assisting and enhancing the surgeon's performance by increasing the autonomy of surgical robots has fundamental importance. The technique of teaching by demonstration (TbD) is capable of transferring the manipulation skills from human to humanoid robots by employing active learning of multiple demonstrated tasks. With the improved ability to interact with humans, such as flexibility and compliance, the new generation of serial robots becomes more and more popular in nonclinical research. Thus, advanced control strategies are required by integrating cognitive functions and learning techniques into the processes of surgical operation between robots, surgeon, and minimally invasive surgery (MIS). In this article, we propose a novel methodology to model the manipulation skill from multiple demonstrations and execute the learned operations in robot-assisted minimally invasive surgery (RA-MIS) by using a decoupled controller to respect the remote center of motion (RCM) constraint exploiting the redundancy of the robot. The developed control scheme has the following functionalities: 1) it enables the 3-D manipulation skill modeling after multiple demonstrations of the surgical tasks in open surgery by integrating dynamic time warping (DTW) and Gaussian mixture model (GMM)-based dynamic movement primitive (DMP) and 2) it maintains the RCM constraint in a smaller safe area while performing the learned operation in RA-MIS. The developed control strategy can also be potentially used in other industrial applications with a similar scenario.

Robot-Assisted systems for Minimally Invasive Surgery enhance the surgeon capability, however, direct control over both the surgical tools and the endoscope results in an increased workload that leads to longer operation times. This work investigates the introduction of SCAN (System for Camera Autonomous Navigation) to overcome this limitation. An experimental study involving 12 participants was carried out with the da Vinci Research Kit. Each user tested two novel camera control modalities, autonomous and semi-autonomous, as well as the current manual control of the camera, while carrying out a dry-lab task. Among the camera control modalities, the autonomous navigation achieved better objective performances and the highest user confidence. Moreover, the autonomous control (along with the semi-autonomous one) was able to optimize some metrics related to the robotic surgery workflow.

Hands-on training is an indispensable part of surgical practice. As the tools used in the operating room become more intricate, the demand for efficient training methods increases. This work proposes a robotic assistance-as-needed method for training with surgical teleoperated robots. The method adapts the intensity of the assistance according to the trainee's current and past performance while gradually increasing the level of control of the trainee as the training progresses. The work includes an experiment comprising 160 acquisition sessions from 16 novice subjects performing a bimanual teleoperated exercise with a da Vinci Research Kit surgical console. Results capture the subtleties in the task's learning curve with and without robotic assistance and hint at the potential of robotic assistance for complex visuomotor training. Although robotic assistance for motor learning has received mixed results that range from beneficial to detrimental effects, this study shows such assistance may increase the rate of learning of certain skills in complex motor tasks.

The increasing availability of inexpensive software defined radios (SDRs) allows nowadays to implement cognitive radio (CR) functionalities in large scale networks such as the Internet-of-Things and future 5G systems. In this work, we focus on the spectrum sensing functionality that must take into account the front-end impairments of low-cost devices. Based on the noise model of a real SDR dongle, we address the problem of robust signal detection in the presence of noise power uncertainty and non-flat noise power spectral density (PSD). In particular, we analyze the receiver operating characteristic (ROC) of different known detectors in the presence of such front-end impairments, to understand the performance attainable in a real-world scenario. Based on the analysis, we propose two frequency-domain detectors that are proven to outperform previously proposed spectrum sensing techniques such as, e.g., eigenvalues-based tests.

A common approach for estimating the number of wireless sources is to adopt model order selection based on information theoretic criteria (ITC). In this paper we study the generalized information criterion (GIC) and propose a design method for setting the penalty in practical situations, where the sample size is finite. The design is based on the maximum probability of correct model selection, that can be approximated using the statistic of the ratio between the largest eigenvalue and the trace of a white central Wishart matrix. For this metric we provide the exact distribution and a new approximation.

Spectrum sensing is an essential functionality in cognitive radio (CR) systems allowing us to discover spectrum opportunities and enabling primary user (PU) protection. Wideband spectrum sensing (WS) improves the awareness of the surrounding radio environment by jointly monitoring multiple frequency bands. In this paper we propose a WS approach based on the observation of a frequency domain representation of the received signal and the adoption of model order selection (MOS) to identify the occupied frequency components. We provide a general formulation of the problem valid for any kind of spectral representation and then focus on the case in which discrete Fourier transform (DFT) is used. This choice is motivated by the fact that DFT blocks are available in many wireless systems, such as OFDM receivers and recently proposed software radio architectures. We provide analytical expressions for the maximum probability of correct selection of the occupied sub-bands valid for MOS approaches encompassed within the generalized information criterion (GIC). We then propose a method for designing the selection algorithm to balance overestimation and underestimation. Numerical results show that the MOS scheme derived for DFT can be successfully applied also when more accurate frequency representations, such as multitaper (MT) spectrum estimates, are adopted.

In this paper we present a thorough study of spectrum sensing performance in cognitive radio (CR) scenarios where the primary user (PU) transmission is not continuous. In particular, we consider a sensing scheme in which the spectrum is monitored periodically for a fraction of time. In such a situation, sensing is affected by common detection impairments, including noise and fading, as well as by the PU temporal behavior. It is thus necessary to properly design periodic sensing parameters to balance between sensing overhead and detection performance. In this context, we derive a comprehensive analytical framework which accounts for detector performance, presence of noise and fading, PU temporal statistics, and periodic sensing. The analysis allows expression of the detection and false alarm probabilities in closed-forms to capture an explicit relationship between the PU temporal statistic and periodic sensing parameters. Our results show that the temporal behavior of the PU have a significant impact on the detection performance, and therefore a proper design of the sensing parameters is important. Based on our analysis we propose useful strategies for the design of effective periodic sensing.

Wideband spectrum sensing (SS) allows cognitive radios (CRs) to reach, by monitoring large portions of spectrum, a better awareness of the surrounding radio environment. In this paper, we formulate wideband SS as a model order selection problem. This approach consists in the adoption of information theoretic criteria (ITC) to identify the occupied frequency components in a frequency domain representation of the observed signal. We provide a general formulation of the problem and then focus on the case in which discrete Fourier transform (DFT) is used as spectral representation. Finally, we propose consistent ITC for which we provide analytical expressions for the maximum probability of detection.

In cooperative spectrum sensing networks the sensing nodes are often assumed to have the same noise power level. However, the different secondary users (SUs) could experience different temperatures, have receiver chains with different characteristics or even with a completely different architecture. Therefore, it is unlikely that they experience exactly the same noise power. In this paper we study the problem of cooperative spectrum sensing in cognitive radio (CR) networks, focusing on the case where the receivers experience different levels of noise power (uncalibrated receivers). We propose the independence test and compare it with the popular sphericity test. The independence test is in fact the generalized likelihood ratio (GLR) when the SUs are uncalibrated. We address in particular the threshold setting problem under a Neyman-Pearson framework. In order to reduce the complexity of the analysis, we approximate the test metrics as beta distributed random variables (r.v.s), by using a moment-matching approach. We provide simple and analytically tractable expressions for the computation of the probability of false alarm and for setting the decision threshold. Numerical simulations show that these approximated forms match very well the empirical distributions, allowing a very precise estimation of the probability of false alarm. In cooperative spectrum sensing with uncalibrated receivers the independence test is shown to be robust to strong imbalances of the noise power level.

In this work we perform an asymptotic analysis of estimated noise power (ENP) energy detector (ED) to derive the condition for the existence of the SNR wall phenomenon. We prove that an ED with noise estimation does not exhibit the SNR wall if the variance of the estimate reduces when the observation time increases. In the absence of SNR wall, we show that the maximum slope of the design curves (SNR vs. observation time for an arbitrary target probability of false alarm (Pfa) and probability of detection (Pd)), equal to -5 dB/decade for the ideal ED, can be reached also by an ENP-ED. Finally, we derive analytical expressions for the design curves when maximum likelihood (ML) noise power estimation is adopted, and we prove that, asymptotically, the signal-to-noise ratio (SNR) penalty with respect to ideal ED is of 1.5 dB, when the number of noise-only samples is equal to the number of observed samples.

An uncertain knowledge of the noise power level can severely limit the energy detector (ED) spectrum sensing capability. In some situations this uncertainty can cause signal-to-noise ratio (SNR) penalties or even the rise of the SNR wall phenomenon. In this paper we analyze the performance of the ED with estimated noise power (ENP), addressing the threshold design and giving the conditions for the existence of the SNR wall. We derive analytical expressions for the design curves (SNR vs. observation time for a target performance) for the ENP-ED. Then we apply our analysis to cognitive radio (CR) systems where energy detection is used for fast sensing. For example it is shown that the SNR penalty with respect to ideal ED is of 5 log10(1+λ/λ) dB, when the time dedicated to noise power estimation is a multiple λ of the ED observation interval.