Sensorimotor synchronisation

We consider a signaling format where the information to be communicated from one or multiple transmitters to a receiver is modulated via a superposition of independent data streams. Each data stream is formed by error-correction encoding, constellation mapping, replication and permutation of symbols, and application of signature sequences. The relations between the data bits and modulation symbols transmitted over the channel can be represented by a sparse graph. In the case where the modulated data streams are transmitted with time offsets the receiver observes spatial coupling of the individual graphs into a graph chain enabling efficient demodulation/decoding. We prove that a twostage demodulation/decoding method, in which iterative demodulation based on symbol estimation and interference cancellation is followed by parallel error correction decoding, achieves capacity on the additive white Gaussian noise (AWGN) channel asymptotically. We compare the performance of the two-stage receiver to the receiver which utilizes hard feedback between the error-correction encoders and the iterative demodulator.

The increased number of cores integrated on a chip has brought about a number of challenges. Concerns about the scalability of cache coherence protocols have urged both researchers and practitioners to explore alternative programming models, where cache coherence is not a given. Message passing, traditionally used in distributed systems, has surfaced as an appealing alternative to shared memory, commonly used in multiprocessor systems. In this thesis, we study how basic communication and synchronization primitives on manycore processors can be improved, with an accent on taking advantage of message passing. We do this in two different contexts: (i) message passing is the only means of communication and (ii) it coexists with traditional cache-coherent shared memory. In the first part of the thesis, we analytically and experimentally study collective communication on a message-passing manycore processor. First, we devise broadcast algorithms for the Intel SCC, an experimental manycore platform without coherent caches. Our ideas are captured by OC-BCAST (on-chip broadcast), a tree-based broadcast algorithm. Two versions of OC-BCAST are presented: One for synchronous communication, suitable for use in high-performance libraries implementing the Message Passing Interface (MPI), and another for asynchronous communication, for use in distributed algorithms and general-purpose software. Both OC-BCAST flavors are based on one-sided communication and significantly outperform (by up to 3x) state-of-the-art two-sided algorithms. Next, we conceive an analytical communication model for the SCC. By expressing the latency and throughput of different broadcast algorithms through this model, we reveal that the advantage of OC-BCAST comes from greatly reducing the number of off-chip memory accesses on the critical path. The second part of the thesis focuses on lock-based synchronization. We start by introducing the concept of hybrid mutual exclusion algorithms, which rely both on cache-coherent shared memory and message passing. The hybrid algorithms we present, HYBLOCK and HYBCOMB, are shown to significantly outperform (by even 4x) their shared-memory-only counterparts, when used to implement concurrent counters, stacks and queues on a hybrid Tilera TILE-Gx processor. The advantage of our hybrid algorithms comes from the fact that their most critical parts rely on message passing, thereby avoiding the overhead of the cache coherence protocol. Still, we take advantage of shared memory, as shared state makes the implementation of certain mechanisms much more straightforward. Next, we try to profit from these insights even on processors without hardware support for message passing. Taking two classic x86 vii Preface processors from Intel and AMD, we come up with cache-aware optimizations that improve the performance of executing contended critical sections by as much as 6x.

The methods of fractional calculus are nowadays shown to have wide range of applications in various branches of applied physics, especially in control theory, viscoelasticity, diffusion, turbulence, electrodynamics, signal processing, etc. This paper is devoted to a numerical analysis of different computational methods applied to nonlinear fractional differential equations, in particular, to fractional-order Lorenz system. The system belongs to a class of nonlinear systems exhibiting doublescroll chaotic attractors at some parameter values, and it is often used by engineers, physicists and mathematicians as a toy-model exhibiting similar features with real physical systems. Numerical solutions are provided by using two different computational methods: Multistep Differential Transformation method and Multistep Differential Transformation method with drops, where the fractional derivatives are defined in the sense of Caputo. We quantify the distinction between the integration methods ...

Is the information transmitted by an ensemble of neurons determined solely by the number of spikes "red by each cell, or do correlations in the emission of action potentials also play a signi"cant role? We derive a simple formula which enables this question to be answered rigorously for short timescales. Using a power series expansion, the mutual information the ensemble of neurons conveys about external stimuli is broken down into "ring rate and correlation components. This analysis leads to a new quanti"cation procedure which is directly applicable to simultaneous multiple neuron recordings. It also shows rigorously that even when noise correlations are not stimulus modulated, synergistic e!ects can arise if signal and noise correlation are of opposite sign.

Despite its obvious suitability for distributed multimedia applications, multicasting has not yet found widespread application. Having analyzed shortcomings of today's approaches, we devise in the GCAP project a new endto-end transport architecture for multimedia multicasting that supports partial order and partial reliability. In this paper, we argue that, at the network layer, single-source multicasting (PIM-SSM) should be chosen. Consequently, our Monomedia Multicast protocol provides, along with reliability and QoS monitoring functionality, an ALM based multicast solution referred to as TBCP (Tree Building Control Protocol), to be used as back channel for SSM, e.g. for retransmission requests. On top of the Monomedia protocol, our Multimedia Multicast protocol handles multimedia sessions composed of multiple monomedia connections: The FPTP (Fully Programmable Transport Protocol) allows applications to specify, through its API, the (global) synchronization and (individual) reliability requirements within a multimedia session. Our group management approach is focused on group integrity.

Most of the standard receivers used for the common-view time transfer are of the single-channel, singleaequency, Coarse/Acquisition code type, They all are built according to the "NBS standard" prototype from the early 1980's. The uncertainty of the commonview time comparisons varies between 2-3 ns RMS for the best of them. The new generation of the unexpensivc, multi-channel C/A code receivers which are now available seems to make the old standard obsolete. One of the interesting new solutions, proposed for the time transfer, is the 8-channel Motorola VP Oncore receiver combined with a time-interval counter and a microcomputer (Gifford et al., 1996), (Lewandowski et al ., 1996). Such a measurement system was tested for several months at the Time Section of the Bureau International des Poids et Mesures (BIPM) in Skvres, yelding very interesting results (Lewandowski et al., 1997). To maintain the compatibility with the old "NBS type'' receivers, still in use at the majority of time laboratories, and to make use of the Motorola multi-channel capabilities, special software driving the receiver and the counter was developed at BIPM in 1997. The results of tests in time transfer between BIPM in Skvres and AOS in Borowiec (Astrogeodynamical Observatory in Borowiec, Poland) at the baseline of about 1200 km give the uncertainty of results in the range of 1-2 ns R M S when applying recommendations of CCDS (Allan, Thomas, 1993).

The airborne F-SAR sensor, developed and operated by the DLR Microwaves and Radar Institute, provides very high resolution, fully polarimetric synthetic aperture radar (SAR) data in multiple frequency bands that are enhanced by several interferometric imaging modes. The high resolution/high bandwidth regime of operation introduces significant challenges regarding both data calibration and image formation, or focusing. This paper reviews the methods for accurate antenna characterization and the parameter extraction for SAR data processing. The antenna measurements must accurately portray interactions with the surroundings as mounted on the aircraft. The results presented concern the antenna characterization itself and illustrate how the quality of real polarimetric and interferometric SAR data is significantly improved. Antenna pattern characterization, Compact Test Range, SAR processing I.

The Microwaves and Radar Institute of the German Aerospace Center (DLR) is known for its consistent work on the field of airborne synthetic aperture radar and its application. In April 2008 we celebrate the 20 th anniversary of the maiden flight of the well-known E-SAR system. E-SAR was maintained well over the time. It provided valuable knowledge to the science community, especially in the last 10 years. However, it became more and more obvious that a technological renewal was inevitable. Consequently the development of a new SAR system was put on the line some years ago under the name 'F-SAR'.

This paper presents a new setting free approach to synchronization of two-end current and voltage unsynchronized measurements. Authors propose new non-iterative algorithm processing three-phase currents and voltages measured under normal steady state load condition of overhead line. Using such synchronization line parameters can be estimated and then precise fault location can be performed. The presented algorithm has been tested with

This paper presents a novel fuzzy logic based control method for shunt active filters. Since the fuzzy sets are based on linguistic description, therefore they don't need to the mathematical model of the investigated systems. The proposed method is very suitable to nonlinear and time variant loads. The controller is robust, reliable and it has a smooth response. Also transient response of method is much better than the other classical methods. The simulation results confirm the suitable performance of the filter using this control method.

Aims and Scope: The Advances in Computer Science and Engineering is a peer-reviewed international journai. which publishes originai research papers and criticai sulvey articies in aii acpects of ComputeiScience and Engineeting. Experimental papeis ere also published. The ACSEis pubiished in two volumss annuaiiy aod each volume compees of fwo issues appearing in Februury, May. August and November lndexing and Reviewe: Ali published articles are reviewed in Zentralblatl f"t Mathematik, Indexcopemicus Data and the ACSE has been evaluated for inclusian in Scopus, Elseviei's bibliographic database, Ei databases index, EMBASE, EMCare aisa. Submission of Manuscnpts: Authors may submit their papers for wnsideration in the ACSE by the following modes: 1. Online submission (only .pdf and .doc fites): Piease uisit journai's homepage at hHp:l/www.pphmj.cam/joumalYacse.htm 2. Electronically (only .tex, .dvi,. p d t .ps and .do= iites

Frequency and phasor measurement of time-varying signals has assumed great importance for phasor measurement units for wide area power networks in micro and smart grid environments. It is well known that the power signals undergo dynamic transitions during power quality and fault type disturbances. Also accurate estimation of amplitude, phase and frequency of a sinusoid in the presence of harmonics/inter harmonics and noise plays an important role in wide variety of power system applications, like protection, control and state monitoring. One of the widely used approach like the Fourier linear combiners becomes ineffective to track the sudden changes in the power signal amplitude, phase, and frequency in the presence noise and distortions. This paper, therefore, presents a new adaptive filter for estimating the frequency and phasors of sinusoids mixed with decaying dc and harmonic components undergoing sudden transitions. The filter is based on a quadratic polynomial signal model whose parameters are estimated using a variable step continuous p-norm LMS algorithm. Extensive computer simulations have been carried out with synthetic signals to demonstrate the viability of this new approach. Further signals obtained from a distributed generation system in MATLAB/SIMULINK environment are used for frequency, fundamental and harmonic phasor estimation.

A conventional receiver performs timing recovery and equalization separately. Specifically, conventional timing recovery is based on a phase-locked loop that relies on the decision provided by its own symbol detector. We propose a new timing recovery scheme based on per-survivor processing (PSP) that jointly performs timing recovery and equalization for uncoded partial response channels. In the proposed scheme, each survivor of the Viterbi algorithm maintains its own estimate of the timing offset, and this estimate is updated according to the history data associated with the survivor path. As compared to conventional timing recovery at BER = 10 −4 , the proposed scheme can provide a 0.5 dB gain in SNR.

When realizing carrier frequency offset estimation at low signal-to-noise ratios, a typical feed-forward synchronization unit solely relies on known pilot symbols. The importance of the position of these pilot symbols within the burst has been elaborated on in the literature. In this paper, we discuss the importance of the pilot symbol constellation for iterative code-aided synchronization. Furthermore, we investigate the transmission order of data symbols for a turbo coded system with code-aided synchronization.

SpiNNaker (a contraction of Spiking Neural Network Architecture) is a million-core computing engine whose flagship goal is to be able to simulate the behaviour of aggregates of up to a billion neurons in real time. It consists of an array of ARM9 cores, communicating via packets carried by a custom interconnect fabric. The packets are small (40 or 72 bits), and their transmission is brokered entirely by hardware, giving the overall engine an extremely high bisection bandwidth of over 5 billion packets/s. Three of the principle axioms of parallel machine design-memory coherence, synchronicity and determinismhave been discarded in the design without, surprisingly, compromising the ability to perform meaningful computations. A further attribute of the system is the acknowledgment, from the initial design stages, that the sheer size of the implementation will make component failures an inevitable aspect of day-today operation, and fault detection and recovery mechanisms have been built into the system at many levels of abstraction. This paper describes the architecture of the machine and outlines the underlying design philosophy; software and applications are to be described in detail elsewhere, and only introduced in passing here as necessary to illuminate the description.

Clock synchronization protocols for packet-oriented networks, like IEEE 1588, depend on time stamps drawn from a local clock at distinct points in time. Due to the fact that software-generated time stamps suffer from jitter caused by nondeterministic execution times, many implementations for high precision clock synchronization rely on hardware support. This allows time readings for packets with very low jitter close to the physical layer. Nevertheless, approaches using hardware support have to carefully consider influences on synchronization accuracy when it comes to the range of nanoseconds. Among others, limits come from the update interval, oscillator stability, or hardware clock frequency. This paper enlightens the limits for such implementations based on an analysis of the influences of the main factors for jitter. The conclusions give hints for efficiently optimizing current implementations.

Interpersonal synchronization is fundamental for motor coordination during social interactions. Discerning emergent (entrainment) from planned synchronization represents a non-trivial issue in visually bonded individuals acting together, as well as assessing whether inter-individual differences, e.g., in autistic traits, modulate both types of synchronization. In a visuomotor finger-tapping task, two participants replicated a target tempo either synchronizing ('joint' condition) or not ('non-interactive' condition, 'non-int') with each other. One participant was exposed ('induced') to tempo aftereffect (a medium tempo seems faster or slower after exposure to slower or faster inducing tempi), but not the other participant ('not induced'); thus they had different timing perceptions of the same target. We assessed to what degree emergent and/or planned synchronization affected dyads by analyzing inter-tap-intervals, synchronization indexes, and cross-correlation coefficients. Results revealed a 'tempo aftereffect contagion': inter-tap-intervals of both induced and not-induced participants showed aftereffect in both the joint and non-int conditions. Moreover, aftereffects did not correlate across conditions suggesting they might be due to (at least in part) different processes, but the propensity for tempo aftereffect contagion correlated with individuals' autistic traits only in the non-int condition. Finally, participants coadjusted their tapping more in the joint condition than in the non-int one, as confirmed by higher synchronization indexes and the mutual adaptation pattern of cross-correlation coefficients. Altogether, these results show the subtle interplay between emergent and planned interpersonal synchronization mechanisms that act on a millisecond timescale independently from synching or not with the partner.

This paper presents a new frequency-adaptive synchronization method for grid-connected power converters which allows estimating not only the positive-and negativesequence components of the power signal at the fundamental frequency, but also other sequence components at multiple frequencies. The proposed system is called the MSOGI-FLL since it is based on a decoupled network consisting of multiple second order generalized integrators (MSOGI) which are frequency-adaptive by using a frequency-locked loop (FLL). In this paper, the MSOGI-FLL is analyzed and its performance is evaluated by both simulations and experiments.

Grid synchronization algorithms are of great importance in the control of grid-connected power converters, as fast and accurate detection of the grid voltage parameters is crucial in order to implement stable control strategies under generic grid conditions. This paper presents a new grid synchronization method for three-phase three-wire networks, namely dual second-order generalized integrator (SOGI) frequency-locked loop. The method is based on two adaptive filters, implemented by using a SOGI on the stationary αβ reference frame, and it is able to perform an excellent estimation of the instantaneous symmetrical components of the grid voltage under unbalanced and distorted grid conditions. This paper analyzes the performance of the proposed synchronization method including different design issues. Moreover, the behavior of the method for synchronizing with highly unbalanced grid is proven by means of simulation and experimental results, demonstrating its excellent performance.

Cortical synchronization in the gamma-frequency range (above ~30.0 Hz) and the signal/noise interplay described by stochastic resonance models have been proposed as basic mechanisms in neuronal synchronization and sensory information processing, particularly in vision. Here we report an observation in humans of linear and inverted-U distributions of the electrophysiological (EEG) responses to visual contrast stimulation in the gamma band and in the low frequency components of the visual evoked responses (VER), respectively. The combination of linear and inverted-U distributions is described by a stochastic resonance model (SR). The observation needs replication in larger subjects’ samples. It nevertheless adds to the available evidence of a role of gamma oscillatory signals and SR mechanisms in neuronal synchronization and visual processing. Some functional adaptation in human vision appears conceivable and further investigation is warranted.

ONERA experimented bistatic satellite to aircraft SAR imaging with its lightweight SAR system BuSARd receiving TerraSAR transmitted signal at X-band with two channels for direct and terrain scattered paths. Static receiver experiment shown that the direct path signal is instrumental in clock synchronisation. Later, two flight experiments with anti-parallel and parallel paths were conducted showing that the trajectory error influences the focusing mainly through the clock synchronisation error from non-modeled direct path length. Significance of the receiving aircraft velocity was also emphasized from the anti-parallel acquisition.

Efficiently transferring information, which is of major importance for almost all engineering systems, requires a data representative appropriate to maintain the density and consistency of the information dispatched. As such a machine-readable representation of data, barcodes are one of the most recognized and powerful instruments of the purpose in every domain. Likewise, from the perspective of engineering, colors are excellent sources of information exchange, since the transmission of color connotes the conveyance of its entire scalar attributes in the same spatial channel component. Exerting colors on barcodes as an effective way of bursting the data conveyance capacity has been an active area of research for over 50 years. Significant progress has been achieved through efforts in this regard. It is also envisagable that the evident evolution in related technologies exoterically empowers the enhancement of color barcode capabilities as capacity and reliability, thereby further encouraging prospected research in this direction. Herein, a comprehensive survey of the studies on this main area of interest is presented. To help better acquainted with the field, also a taxonomy of the peculiar interference sources and distortion effects is provided, besides, the 3D barcoding process itself and the research areas are described. Most of the relevant works from debut to the present are broadly examined. Rather than presenting a timeline, studies that pertain to similar issues are addressed together. Amongst those related, premising or pivotal ones are preferably cited as far as feasible. Moreover, all perused works are analyzed by the research areas and the results are presented. Also, the issues relatively more prominent as affecting the performance of the whole process are specified. In the conclusion, some of the research subjects that appear open, scarce, or require further elaboration were remarked on as well. It is anticipated this study to contribute to the efforts toward leveraging color in barcodes.

Robust image watermarking algorithms have been proposed among others as methods for discouraging illicit copying and distribution of copyright material. Having robustness to pixels modifications in mind, many watermarking designers use techniques coming from the communications domain such as spreading spectrum to embed hidden information, be it in the spatial or in the transform domain. Most of the attacks dedicated to make watermarking algorithms inefficient degrade images through geometric distortions. One solution to counter them is to add synchronization information. In this paper we present an analysis of this type of distortions and we propose a metric to estimate the distortion undergone by an image. This metric is content independent, invariant to global translation, rotation and scaling, which can be considered as non-meaningful transformations. To demonstrate the relevance of this metric, we compare some of its results with the subjective degradation of the image produced by the Stirmark software.

This paper considers the design of multi-objective iterative learning control (ILC) schemes for discrete linear systems. In particular, a two step design approach is developed where the feedback and learning controllers are designed separately. The design procedures developed are able to impose performance specifications over selected frequency ranges to obtain the desired shapes of the sensitivity functions relevant to a particular example. This makes it possible to reject disturbances in specific frequency ranges and hence the desired performance specifications for both the transient response and error convergence speed can be achieved. All design computations required can be completed using linear matrix inequalities (LMIs). Preliminary experimental results demonstrate the effectiveness of the new design.

ONERA experimented bistatic satellite to aircraft SAR imaging with its lightweight SAR system BuSARd receiving TerraSAR transmitted signal at X-band with two channels for direct and terrain scattered paths. Static receiver experiment shown that the direct path signal is instrumental in clock synchronisation. Later, two flight experiments with anti-parallel and parallel paths were conducted showing that the trajectory error influences the focusing mainly through the clock synchronisation error from non-modeled direct path length. Significance of the receiving aircraft velocity was also emphasized from the anti-parallel acquisition.

The paper is devoted to experimental study of possibilities for consensus control of the unbalanced rotors for the machine used in the vibration technology. This problem is treated as an ensuring the prescribed phase shift between the vibrating rotor actuators and has a significant value for the various kinds of the vibration technology, such as vibrating transportation, drilling, and so on. It can not be considered without taking into account the natural tendency of in-phase or anti-phase synchronization of the mechanically coupled oscillating units (cf. Huygens clock synchronization). This phenomenon prevents the controller efforts in maintaining the desired value of the phase shift. In the paper, the control laws for frequency stabilization simultaneously with cooperative control for assuring the desired phase shift between the rotors angular positions are derived and experimentally studied by means of the mechatronic vibration testbed. It is obtained that for the low and medium frequencies the self-synchronization of unbalanced rotors does not prevent achieving the desired phase shift between the rotors. For a high frequency band, the Huygens self-synchronization of rotors manifests itself, narrowing the range of the achievable phase shift.

This contribution describes a distributed multi-camera capture and processing system for real-time media production applications. Its main design purpose is to allow prototyping of distributed processing algorithms for free-viewpoint applications, but the concept can be adapted to other (multicamera) applications. The system integrates broadcast components into a distributed IT-based processing system. The problems that are addressed in this contribution are the synchronisation of these subsystems and the data and control flow. For the synchronisation of the broadcast and IT components we developed a time stamp mechanism based on tightly synchronised PC clocks. This time is used as the master clock of the system. The video frames from the gen-locked multi-camera streams are implicitly synchronised by adjusting their relative timing against this system clock. For the purpose of synchronising the PC clocks two protocols (NTP and PTP) were evaluated. The second main contribution is the processing of multi-camera data in the distributed system. For this purpose a software framework was developed based on a distributed concurrency system implemented in Python.

The infinite-server queueing models M|G|∞, BM|G|∞, BMk|Gk|∞ with homogeneous and non-homogeneous arrivals of customers and catastrophes are considered. The probability generating functions (PGF) of joint distributions of numbers of busy servers and different type served customers, as well as the Laplace-Stieltjes Transforms (LST) of distribution of busy period and distribution of busy cycle for the models are found.

The infinite-server queueing models M|G|∞, BM|G|∞, BMk|Gk|∞ with homogeneous and non-homogeneous arrivals of customers and catastrophes are considered. The probability generating functions (PGF) of joint distributions of numbers of busy servers and different type served customers, as well as the Laplace-Stieltjes Transforms (LST) of distribution of busy period and distribution of busy cycle for the models are found.

The vehicle routing problem with route balancing is a bi-objective routing problem, in which the total route length and the balance of routes (i.e. the difference between the maximal and minimal route length) have to be minimized. In this paper, we propose an approach based on two solution representations: a giant tour representing a sequence of customers (indirect representation) and a complete solution with a decomposition of the giant tour, combined with a split algorithm to alternate between them. This approach offers a mainly efficient way to explore the solution space. Our motivation is based on the possibility to generate efficiently several solutions a time using an indirect representation which has been already proved to be efficient in numerous routing problems resolution. The originality here is to tune the split algorithm considering two objectives. An evolutionary path relinking algorithm is embedded to improve the obtained solutions. The proposed approach is evaluated on classical vehicle routing problem instances and the results push us into accepting that the method is competitive with the best published mono-objective methods (on criteria one : the total route length). On a bi-objective point of view, our method is competitive with the lexicographic solutions reported in the literature in the sense that it provides similar or better results in comparable computational time.

Pausible clocking schemes have been proposed by GALS architects as a promising mechanism for reliable data transfer between synchronous modules fed by low-speed independent clocks. In this paper, we argue that existing schemes are not well-suited for interfacing high-speed IP cores with large clock-distribution tree delay and high communication rates. We propose an alternative interface circuit design for such IP cores that works with partial handshake between communicating modules and minimizes the performance penalty of the sender and receiver. Our circuit, unlike pausible clocking, has a small probability of failure.

The Controller Area Network with Flexible Data-Rate (CAN FD) is the new generation of CAN technology to meet the daily increasing bandwidth requirement for Automotive Cyber-Physical Systems (ACPS). However, ACPS is a securitycritical system, an efficient security/timing-aware Design Space Exploration (DSE) method is required to fully utilize CAN FD's high data phase data-rate. In this paper, we propose an AUTOSAR-compliant system model that integrates both timing and security constraint, an integrated Mixed-Integer Linear Programming formulation (i-MILP) for the optimal DSE of CAN FD, and a divide-and-conquer approach to the i-MILP (dc-MILP) to address its timing complexity problem. The experiment results show that dc-MILP scales well for industrial-size systems, and saves 1.94% to 4.76% bandwidth utilization and guarantees the schedulability for more signal sets by comparing with state-ofthe-art algorithm.

For two positive maps φ i : B(K i) → B(H i), i = 1, 2, we construct a new linear map φ : B(H) → B(K), where K = K 1 ⊕ K 2 ⊕ C, H = H 1 ⊕ H 2 ⊕ C, by means of some additional ingredients such as operators and functionals. We call it a merging of maps φ 1 and φ 2. The properties of this construction are discussed. In particular, conditions for positivity of φ, as well as for 2-positivity, complete positivity, optimality and nondecomposability, are provided. In particular, we show that for a pair composed of 2-positive and 2-copositive maps, there is a nondecomposable merging of them. One of our main results asserts, that for a canonical merging of a pair composed of completely positive and completely copositive extremal maps, their canonical merging is an exposed positive map. This result provides a wide class of new examples of exposed positive maps. As an application, new examples of entangled PPT states are described.