A novel condition for complex FIR digital filters to have linear phase is presented. It naturally includes the linear phase conditions for real FIR filters as special cases.

In this paper, we show that if the spectrum of a frequency modulated (FM) signal is discrete (generally nonuniformly spaced) then the modulating signal is also periodic and-generally-nonuniformly spaced in the frequency demain. In terms of mathematics if an FM signal is almost periodic (a.p.), then the modulating signal is also a.p.

Mode partition noise of a 1.55 µm DFB laser is suppressed by an optical injection locking technique for the first time. The effectiveness is confirmed in a 2.4 Gbit/s, 40 km transmission experiment. This technique completely eleminates the bit-error-rate floor of 10^-4.

This letter describes a method of estimating the 0 isotherm height in rain, based on analyses of temperature profiles in rain and the Aerological Data of Japan. The derived formula agrees fairly well with the provisional one in the CCIR prediction method of rain attenuation on Earth-space paths.

In many occasions, radio-relay system planners need the data of radio-relay antenna beam bending due to atmospheric refraction. The authors have derived approximate numerical formulas for such atmospheric bending which will facilitate computer calculations.

The original iterated codes proposed by P. Elias can be regarded as the codes constructed by iterating two dimensional product codes. While the modified product codes have been proposed and shown to be able to increase the rates without increasing the probability of decoding error. On the other hand, we have proposed new codes, called modified iterated codes A, and improved the performance for the original iterated codes by applying the coding and the decoding schemes of the modified product codes to these product codes. It has been proved that the rates of codes A were always much higher than those of the original iterated codes for cross-over probability p0.1617. In this paper, by additionally combining the coding and the decoding schemes of the superimposed codes constructed on the basis of the modified product codes, which are able to increase the rates with the same minimum distance, the performance for codes A can be further improved. We call codes thus constructed modified iterated codes B. It is shown that the rates of codes B are partially higher than those of codes A for p0.0959.

Phase-Locked Loops (PLL's) have been playing an important role in communication systems. In recent years efforts have been shifted toward implementation of the PLL's by means of digital circuits and a number of all digital PLL's (DPLL's) have been proposed to solve the problem of stability in the PLL's. One of the major problems of these DPLL's is the requirement of a high frequency local clock for a good phase lock precision, which inevitably makes it difficult to apply the DPLL's into high frequency operations. In this paper, a DPLL which have a good phase lock precision with a low frequency local clock is proposed. A good phase-lock precision is obtained by small phase control quantum, however, it makes the locking range narrow. Then, frequency control is employed to improve the locking range and a binary quantized phase frequency detector is also described. The relation between clock frequency and performances of the system is analyzed and verified by some experiments. Also, analysis and experimental performance are given for both acquisition behavior and steady-state phase error characteristics with white Gaussian noise present, resulting in that a good phase-lock precision and a wide locking range are obtained with a low frequency clock. The experimental results show a very close agreement with the theoretical results.

The characteristic features of a mutual optical injection-locking system using semiconductor lasers are both analytically and experimentally examined. It is found that in the case of low injection level, locked frequency, locked powers and locking bandwidth of the system are suitably characterized in terms of both a figure of merit specifying unilaterality of optical injection and a modified round-trip phase angle of optical coupling circuit between lasers. The analysis is applicable not only to unilateral optical injection-locking system but also to mutual optical injection-locking system with an arbitrary inverse optical injection. The results of experiment using AlGaAs CSP lasers are found to be in good agreement with the analysis provided that the mutual injection levels are low.

Continuous-time high-frequency active filters suitable for monolithic implementation by standard low cost bipolar process are presented. Balanced NIC's are used to cancel out the loss of passive RC integrators, thus realizing active loss-less integrators with good high-frequency performance. Two types of balanced integrators are proposed and their quality factors are analyzed. The proposed NIC-integrators are used to realize tunable active filters capable of low-voltage operation and without the drawbacks of the NIC-gyrator filters. As an example of application of the NIC-integrators to leapfrog simulation of RLC ladders a second-order 1 MHz bandpass filter was designed, computer simulated and laboratory tested showing good results.

Some classes of hypercube structure and a routing method for them are presented in this paper. The first class of cyclic permutation networks (CPN1) is constructed by connecting input terminals with output terminals for permutation networks in accordance with a given rule. CPN1 is also a cyclic interchange graph of a hypercube, which is constructed by replacing each edge of the hypercube with a vertex and connecting the new adjacent vertices cyclically. The second class of cyclic permutation networks (CPN2) is constructed by splitting each vertex of CPN1 into two new vertices. The CPN2 is also graph-isomorphic to the CCC network. This paper also presents a routing method using the path coefficient for the cyclic permutation networks and the CCC network. For the cyclic networks, the path coefficient which determines the path from any source node (terminal) to the destination node can be expressed using only the binary representation of the destination node, and does not depend on the representation of the source node.

This paper presents an image compression coding sheme named Boundary Detection Coding (BDC). It is worked out to compress digital multi-color images having large uniformly colored areas. Input images to the BDC encoder are given in the raster scan data format. The encoder detects boundary points of uniformly colored areas and changes them into compression codes. It does not generate any code while scanning uniformly colored areas. The BDC decoder changes the boundary codes back to the original image using some characteristics of boundary points. The BDC is much more efficient than the one-dimensional runlenght coding method. This paper covers the coding algorithm of the BDC, the results of some experiments and comparison with other methods.