SHIELDING FLAW DETECTION AND MEASUREMENT IN ORTHOGONAL FREQUENCY DIVISION MULTIPLEXED (OFDM) CABLE TELECOMMUNICATIONS ENVIRONMENT
An egress signal from a broadband communication system (BCS) including orthogonal frequency division multiplex (OFDM) signals is distinguished from noise and broadcast transmissions and authenticated as such by detecting required or optional pilot signals by their frequency, pattern and/or other characteristics such as recurrence rate or cepstral interval. Detection is made more robust by detection of combinations of required or optional pilot tones. Different BCS plants in the same geographic area or overbuilt can be discriminated based on the frequencies and frequency intervals of required, fixed frequency PHY link channel (PLC) related pilots, pattern and frequency of the presence or absence of optional pilots at fixed frequency intervals and/or cepstral or recurrence intervals of required scattered pilots if different symbol rates are employed or frequency of scattered pilots if the OFDM bands do not overlap in the adjacent or overbuilt BCSs.
Latest ComSonics, Inc. Patents:
- Leakage detection for cable TV systems with upstream signals above 118 MHz
- Shielding flaw detection and measurement in quadrature amplitude modulated cable telecommunications environment
- Deriving broadband communication system service area for signal leakage detection
- Miniature mobile marker system and method
- Downstream OFDM signal egress detection
This application is related to U.S. patent application Ser. No. 14/193,076, filed Feb. 28, 2014, which is a division of U.S. patent application Ser. No. 13/080,715, filed Apr. 6, 2011; both of which applications are hereby fully incorporated by reference.
FIELD OF THE INVENTIONThe present invention generally relates to maintenance of cable telecommunication systems and, more particularly, to detection of cable shielding flaws and measurement of signal egress in such systems in which the communicated signal is orthogonal frequency division multiplexed (OFDM).
BACKGROUND OF THE INVENTIONCable telecommunications systems, often referred to as broadband communication systems (BCSs), have been known for a number of years and are currently gaining in popularity and coverage for the distribution of television programming, telephone service and networking of computers such as providing Internet access since they can carry many signals over a wide bandwidth with little, if any interference or distortion, particularly as data transmission rates have increased to accommodate high definition television, increased volume of digital communication and the like. By the same token, since these communications are intended to be confined within the cable system, the increased bandwidth required for such communications need not be allocated from the available bandwidth for other so-called “over-the-air” broadcast communications such as radio, navigation, GPS, emergency communications and the like which must be transmitted as electromagnetic waves through the environment. However, flaws in cable shielding in cable telecommunication systems can allow signal egress which can potentially interfere with broadcast communications and potentially cause hazards. Reciprocally, flaws in cable shielding can permit signal ingress into the cable from the environment and degrade or interfere with the signal being carried by the cable telecommunication system. Therefore, such flaws must be quickly discovered and remedied as they occur due to weather, mechanical damage or the like.
Detection of cable shielding flaws is generally achieved through detection of the signal carried by the cable transmission system that has leaked into the environment, essentially by being broadcast from the shielding flaw. Detection of a signal that has leaked or egressed from a cable flaw may be performed in two stages: first, by a receiver in a mobile vehicle driven in the general vicinity of installed cables that associates a received signal, authenticated as originating from the BCS, with a location of the mobile vehicle using a global positioning system (GPS) receiver which thus reports a general location of a shielding flaw and, second, by a hand-held instrument that can allow repair personnel to follow increasing signal strength to the exact location of the shielding flaw so that repairs and/or maintenance can be carried out.
Of course, such detections must be carried out in an environment in which noise as well as broadcast signals will also be present in the same frequency bands. Accordingly, a problem with all such systems is to identify a received signal as one originating in the cable telecommunications system and numerous techniques have been developed to effectively verify or authenticate a detected signal as a BCS signal. An additional issue that follows from this problem is that a signal which is unique to the cable telecommunication system and distinguishable from broadband noise (e.g. a marker signal) necessarily consumes a finite amount of bandwidth and/or has the potential for interfering with the signal carried by the cable telecommunication system or BCS.
A legacy system seeking to provide a solution to these related issues is disclosed in U.S. Pat. No. 4,072,899, issued Feb. 7, 1978, to Richard L. Shimp, which is hereby fully incorporated by reference. In the system disclosed therein, a variable, relatively low frequency (e.g. “warbled”) audio tone is added as a marker signal to the signal carried by the telecommunication system which can be easily detected by a narrow band portable receiver such that the audio tone can be perceived and followed by maintenance personnel while being easily filtered from or having little effect on the other signals of much higher frequency carried by the cable telecommunication system.
At the present time, the need to carry ever greater amounts of information (e.g. for high definition television (HDTV), digital communications and the like) has resulted in the choice of schemes such as quadrature amplitude modulation (QAM) and orthogonal frequency division multiplexing (OFDM) to increase signals which have increased data content. QAM format signals, as specified in the Data Over Cable Service Interface Specification (DOCSIS) 3.0, can be used in any portion or the entirety of the spectral bandwidth of the BCS system. In general, for QAM signals, a plurality of QAM generators are used, each carrying a small number of channels of information, and their outputs are combined by allocating contiguous spectral bands to each QAM multiplex. The output of a QAM generator or a plurality thereof is often statistically indistinguishable from ambient noise in the environment in which detection must be performed.
OFDM, specified in the recently promulgated DOCSIS 3.1 specification, is a new and, in some aspects, open-ended format proposed for use in the telecommunications industry in which a large number of data or payload subcarriers are grouped into one or more blocks having a bandwidth (in the United States) of 192 MHZ which can be placed at any desired location in the BCS bandwidth. Each of the data subcarriers is of relatively low amplitude (to limit total signal power over the BCS system to levels comparable to earlier signal formats such as QAM) and very narrow bandwidth (e.g. 20 or 40 KHz in the United States depending on the total bandwidth occupied by the OFDM format signal) and spaced at frequency intervals of 25 or 50 KHz equal to the symbol rate and which are quadrature amplitude modulated in accordance with data being transmitted. A number of so-called pilot subcarriers (generally referred to simply as “pilots”) of several different types are also provided which are of slightly greater amplitude which are not modulated by payload data are also provided which may not be modulated but are said to be binary phase shift keyed (BPSK) modulated under the DOCSIS 3.1 specification. The DOCSIS 3.1 specification also provides for so-called scattered pilots (which might be more accurately described as staggered subcarriers but the terminology of DOCSIS 3.1 will be followed) which are systematically shifted in frequency for purposes of signal monitoring (e.g. for adequate signal-to-noise ratio). The specification also suggests, but does not require, additional fixed-frequency pilots at particular frequency intervals. Therefore, an OFDM format signal is generally even less statistically distinguishable from random ambient noise than QAM format signals since its spectral content is essentially flat except for a PHY Link Channel (PLC) exclusion band within an OFDM group. (“PHY” is not an acronym but, rather, an accepted abbreviation of the work “physical” in the context of the physical layer of the seven layer Open Systems Interconnect (OSI) model.)
Accordingly, techniques of signal egress detection for systems prior to that disclosed in the above-incorporated U.S. patent application Ser. No. 14/193,076 directed to egress signal detection in a QAM environment have required the allocation of significant bandwidth (e.g. the equivalent of a band corresponding to a QAM multiplexer or at least the bandwidth corresponding to a television program channel) in order to provide a sufficiently complex signal for detection and identification or authentication of a received signal as a BCS egress signal without causing interference with other information carried by the cable telecommunication system. Allocation of such bandwidth has also been essential to measurement of the strength of signal egress allowing repairs to be prioritized and to assure compliance with regulations governing the operation of BCS plants.
Such an allocation of bandwidth thus reduces the otherwise available bandwidth or data transmission capacity of the cable telecommunication system and is essentially a large fixed cost of operating the system. Even with the allocation of economically significant bandwidth to the shielding flaw detection function, detection is not robust and, where two or more cable telecommunication systems may be present in the same geographic area (referred to as being “over-built”), identification of the individual system having the shielding flaw can often not be performed unless the marker signal is particularly complex; requiring more than minimal bandwidth allocation.
SUMMARY OF THE INVENTIONIt is therefore an object of the present invention to provide a system for detection of egress signals having an OFDM format in accordance with DOCSIS 3.1 and for differentiating between over-built BCS systems without allocation of bandwidth for marker signals or any other distinctive signal other than signals provided in the OFDM signal spectrum, itself.
It is another object of the present invention to facilitate detection and authentication of a received signal as a BCS egress signal and to provide a method and apparatus which exploits mandatory and/or permissive features of an OFDM signal that provides robust detection and authentication of BCS egress signals in a manner that is simplified in comparison with marker signal provision and detection in a QAM environment.
In order to accomplish these and other objects of the invention, a method of detecting an orthogonal frequency division multiplex signal among ambient electromagnetic signals is provided comprising steps of receiving an ambient electromagnetic signal, deriving a spectrum of said ambient electromagnetic noise, and analyzing that spectrum to determine one or more characteristics of patterns of locations or intervals of spectral peaks.
In accordance with another aspect of the invention, a method of authenticating a detected signal as an egress signal from a BCS system is provided wherein a signal carried by said BCS system includes a plurality of pilot or marker signals, said pilot or marker signals being of different types characterized by frequency, pattern of frequency or rate of recurrence, comprising detecting at least two types of said marker or pilot signals having particular frequencies, patterns of frequencies or rates of recurrence.
In accordance with a further aspect of the invention, apparatus for detecting an egress signal from a broadband communication system is provided comprising a tunable receiver to receive ambient electromagnetic signals, a converter for digitizing samples of the ambient electromagnetic signals received by the tunable receiver, and a processor configured to perform computations of a computation of a magnitude squared fast Fourier transform of the digitized samples, a logarithm of results of the magnitude squared fast Fourier transform, and an inverse fast Fourier transform of results of the computation of thee logarithm.
The foregoing and other objects, aspects and advantages will be better understood from the following detailed description of a preferred embodiment of the invention with reference to the drawings, in which:
FIG. 1AA is a high-level block diagram of a shielding flaw detection system as applied to a QAM environment for comparison with
FIG. 1CC is a simulated spectrum of a portion of an OFDM block showing the required pilots near a PLC,
Referring now to the drawings, and more particularly to
It should also be appreciated in this regard, that
The invention disclosed in the above-incorporated application Ser. No. 13/080,715 is directed to detection of marker signals in a QAM signal environment which does not require the dedication of significant bandwidth or any otherwise usable bandwidth to the marker signal. It is possible that an understanding of the present application may be even more readily obtained by comparing and contrasting the present invention with the above-incorporated application as will be discussed in greater detail below, since, in essence, certain aspects of an OFDM format signal and spectrum can be exploited for detection and egress signal authentication in a manner similar to that discussed therein for authenticating a marker signal in a QAM signal environment but with improved robustness and reduced criticality while the marker signal itself need not be separately generated but portions of the OFDM format signal can simply be calibrated to exhibit unique and detectable behaviors.
Referring now to
If a carrier signal of this frequency is then modulated with a signal of varying frequency, two sidebands equally separated in frequency from the carrier will be produced and will exhibit complementary (e.g. opposite) excursions in frequency from the carrier as indicated at (D). This complementary behavior can be detected as will be discussed in further detail below and will be unique to the marker signal. The carrier frequency can then be suppressed in any suitable manner as will be known to those skilled in the art. Further, the complementary behavior of the sidebands avoids any need for a priori knowledge of the signal being carried by the BCS system and, if such a marker signal is placed between QAM bands of different respective frequencies in respective over-built BCS systems, the approximate frequency of the sidebands can allow discrimination between them.
Thus, the marker signal for QAM signal egress detection is developed as a double sideband suppressed carrier signal depicted in
The overall spectrum of OFDM signals in a QAM environment is illustrated in
Referring now to
The frequencies of required PLC-related pilots 230 are used to indicate the location and bandwidth of the exclusion band 260 also referred to as a PHY Link channel (PLC) which is 400 kHz or 800 kHz wide (or 8 or 16 subcarriers at 50 kHz spacing or 16 or 32 subcarriers at 25 kHz spacing). The specified locations of the eight required PLC-related pilots under DOCSIS 3.1 are:
Pilot #1 upper/lower=±550 kHz for a narrow PLC or ±750 kHz for a wide PLC (350 kHz from the PLC edges)
Pilot #2 upper/lower=±1000 kHz for a narrow PLC or ±1200 kHz for a wide PLC (800 kHz from the PLC edges)
Pilot #3 upper/lower=±1,550 kHz for a narrow PLC or ±1,750 kHz for a wide PLC (1,350 kHz from the PLC edges)
Pilot #4 upper/lower=±2,150 kHz for a narrow PLC or ±2,350 kHz for a wide PLC (1950 kHz from the PLC edges).
The pattern of these pilots is easily detectable even if the pilots are BPSK modulated to opposite constellation corners by a pseudo-random sequence running at the OFDM symbol rate, as alluded to above. (In this context, a “constellation” is a diagram representing modulation of a signal by a digital modulation scheme such as QAM or phase shift keying. The constellation displays the signal as a two-dimensional scatter diagram in the complex plane at symbol sampling instants with the entire constellation representing all possible modulation states of the modulated signal. The constellation of BPSK modulation is the simplest possible and has only two possible and symmetrically located states.) The pseudo-random sequence is developed by a 13-bit linear shift register with taps defined by the polynomial x13+x12+x11+x8 (e.g. indicating the shift register stage states fed back to the first shift register stage. In any case, the pattern of pilot subcarriers can be readily detected. The detection of pilots (whether required PLC-related pilots, optional/additional pilots or scattered pilots) is facilitated and simplified since the amplitude is specified to be 6 db above the amplitude of the data subcarriers.
Additional pilots placed at 1.6 MHZ to 6.4 MHZ intervals are suggested but neither the pilots nor the suggested spacing is required by DOCSIS 3.1 and can thus be placed or omitted at will ar any desired frequency or frequencies. These additional or optional pilots may or may not be BPSK modulated, similarly to the required PLC-related pilots discussed above. It should also be appreciated that the pattern of optional pilot subcarriers can also be used to discriminate between overbuilt BCSs as long as the pattern of optional pilots differs in some way (e.g. chosen frequencies, number, separation or the like) between the overbuilt BCSs.
Scattered pilots under the DOCSIS 3.1 specification are always placed at intervals of 128 subcarriers in each symbol (corresponding to an interval of 6,400 kHz for 4 kFFT or 3,200 kHZ for 8 k FFT) and systematically shift in location by 50 kHz (one or two subcarriers) every symbol time (20 μsec. for a 4 k FFT, 40 μsec. for an 8 k FFT) and are thus systematically substituted for data subcarriers as indicated by illustration of OFDM subcarriers with dashed lines in
It should be appreciated that all of the above types of pilots have properties that can be discriminated as being characteristic of an OFDM modulated BCS signal and which, if detected in the environment, are unambiguously indicative of a received egress signal from a BCS. Any or all of the three types of pilots can be detected singly or in any combination for such a purpose.
Additionally, detection is simplified and made more robust since the pilots are required to be approximately 6 db above the amplitude of the data subcarriers. If the same signal power per bandwidth is to be retained from DOCSIS 3.0, each DOCSIS 3.1 subcarrier has an average power level of roughly 25 (for an 8 k FFT) or 50 (for a 4 k FFT)/5,360.537 relative to a 6.0 MHZ DOCSIS 3.0 channel
Returning now to
BCS will now be discussed. The system 100, as illustrated in
Essentially, OFDM processing spreads a substantial number of digital signals over a large number of data subcarriers that are all synchronized and separated in frequency by the symbol transmission rate. Thus each data subcarrier occupies a very narrow frequency band such that 120 or 240 subcarriers are placed in the same bandwidth corresponding to a single QAM band of nominally 6 MHZ bandwidth and each data subcarrier appears as a very narrow (e.g. 25 or 50 KHz) QAM band. The distribution of individual digital data streams in the OFDM subcarriers and the method(s) by which individual digital data streams can be recovered are unimportant to the practice of the invention.
The signal combiner 120 may be any commercially available unit. The output of each of the QAM multiplexers are modulated to be allocated to a particular frequency band (e.g. of 6 MHZ bandwidth for most QAM systems in the United States) but very much smaller data subcarrier bandwidth is employed in an OFDM format signal in which the data or payload subcarriers are similar to very low-bandwidth QAMs) and are independently connected by plural connections 115 and 115′ (only one of which is shown) to signal combiner 120. One or more OFDM signal blocks are also provided to the signal combiner 120, falling within the overall signal spectrum that can be carried by the cable telecommunication system. The output of the signal combiner 120 is then fed to a laser 150 which produces a broadband optical signal that is transmitted over fiber optic link 160 to fiber node 170 where the signal is converted to an electrical signal for distribution over shielded cables 180 to subscribers. It is shielding flaws 190 in cables 180 that the invention is directed to detecting using detector 300.
In accordance with the invention, at the head end 130 of the cable telecommunication system 100, an OFDM signal source 200 may be provided which includes some or all of the pilots described above.
In contrast to the QAM marker signal insertion and detection system described in the above-incorporated patent applications and illustrated in FIG. 1AA, as shown in
As alluded to above, any or all of the pilot subcarriers required or optional under DOCSIS 3.1 or other OFDM specifications that may be foreseeable have characteristics that may be detected in a manner within the level of technology used for detection of added marker signals in a QAM environment but provide more robust and flexible detection because of the increased amplitude of all pilot subcarriers which do not carry the possibility of interference with data transmission. Reliability of detection and authentication of a signal as a BCS egress signal is also enhanced by the number and variety of pilot characteristics and behaviors which can be detected and identified in accordance with the invention. The small difference in detection of continuous pilot subcarriers, whether required of optional, arises from the possibility of BPSK modulation of each continuous pilot subcarrier (whereas each sideband of the QAM marker in the above-incorporated applications is a sinusoid). However, this difference merely affects the shape of the template used for detection and some differences in hardware implementation due to the larger required bandwidth to gather the entire set (e.g. required, optional and scattered) of continuous pilots. The detection of so-called scattered pilot tones includes some substantial differences from detection of QAM marker signals but the unique behaviors of scattered pilot subcarriers also facilitates detection, as will be described below.
Referring now to
It will be recalled from the above discussion that an OFDM signal block is a composite of QAM modulated data subcarriers and pilots of potentially three different types which may or may not be BPSK modulated in accordance with a specified pseudo-random symbol sequence and, even if BPSK modulated, can be treated as single frequency tones although BPSK modulation effectively prevents the pilot subcarrier frequency from being isolated by filtering. On the other hand, BPSK modulation of all pilots in accordance with a pseudo-random sequence causes the frequency shifting due to the BPSK modulation to be exactly synchronized in all pilots which allows the pilots to be identified and distinguished from data sub-carriers (or other non-egress signals that may be received) to detect their presence or absence at particular nominal frequencies (regardless of some variation in amplitude) to a degree entirely sufficient for authentication of a received signal as an egress signal and, therefore, isolation by filtering is not required although it may be convenient if BPSK modulation is not employed.
The data subcarriers 220 (
The required or optional pilots are generated by a pilot signal generator 280 which is depicted as having a separate pilot generator for each type of pilot although, in practice, such separate pilot generators may not be separately implemented. As alluded to above, nominal frequencies of all pilots as well as data subcarriers are related to and derivable from the center frequency or boundary frequencies chosen for the PLC exclusion zone 260 as supplied from a frequency source 290. Further, since the pilots may be BPSK modulated with a pseudo-random sequence operating at the symbol rate, pilot generator 280 is depicted as including a pseudo-random sequence generator, preferably in the form of a linear shift register with feedback from selected stages thereof. The output of pseudo-random sequence generator 282 is provided to the respective generators of each type of pilot subcarriers 284, 296 and 288. The data subcarriers and required and optional pilot subcarriers are then combined in signal combiner 295 to provide the signals in the OFDM block for combination with signal frequencies covering the remainder of the BCS spectrum at signal combiner 120 for transmission over the BCS.
While separate generators are depicted for the three different types of pilots specified by DOCSIS 3.1, the respective pilots would not ordinarily be implemented in such a way due to the difficulty of aligning the respective frequencies in such a possible implementation and the simplicity of an FFT for specification and concurrent generation of all subcarrier frequencies. Rather, the inverse fast Fourier transform (IFFT) of frequencies in the OFDM block spectrum are the frequencies at which both the data subcarriers and pilots are generated.
Referring now to
The preferred processing within the DSP or MCU 350 depicted in a bracket extending therefrom will now be described. Samples of output of the analog-to-digital (A/D) converter 340 are first buffered at 351 and a fast Fourier transform performed on the buffered samples and the magnitude squared of the FFT is computed to approximate the power spectrum as shown at 352. This power spectrum is then cross-correlated (353) with a template (354) which may be generic to, for example, DOCSIS 3.1 signals but is preferably specific to a single BCS of interest. It should be noted that while a template is used in the above-incorporated QAM environment under DOCSIS 3.0, the template used in DOCSIS 3.1 pilot detection is not the same, even if the methods used are similar. The DOCSIS 3.1 template preferably describes an entire set of pilot signals within the received frequency band of an OFDM frequency band, not merely a single marker signal. The maximum point of correlation is then found, at 356. The index at which this maximum occurs is used to create a band-limited sum of the received signal power, as depicted at 355. The maximum correlation value is applied to a threshold detector 357 to determine the presence or absence of one or more egressing OFDM pilots. If it is determined that egress signal leakage is present, the received signal power is the measured leakage level (393). In this latter regard,
This implementation separately tunes a selection of the pilots and places the frequencies into a smaller baseband width in a manner similar to the above-incorporated system for a QAM environment.
Yet another approach could also be achieved by passband sampling (e.g. without an RF down converter) at different rates to alias the selected pilot frequencies into the same smaller bandwidth and using standard techniques to decimate and filter those frequencies to reduce the final sampling rate.
In regard to optional or arbitrary continuous pilot subcarrier, the same techniques described above can be used. The pattern and frequencies of these optional or arbitrary subcarriers is of interest and may (and preferably will) differ between overbuilt BCSs and, if so, would be used to discriminate between them. Similarly, the location in frequency of the PLC and fixed pattern(s) of PLC-related required pilots may differ between overbuilt BCSs and, if so, would be used to discriminate between them.
In general, it is in the interest of BCS operators to use the available choice of center frequency of the PLC exclusion band and the frequencies of optional pilots (and selectable pattern thereof to be unique or at least highly distinct from other BCSs, particularly in the same geographic area so that discrimination between overbuilt or nearby BCSs is possible. That is, detection of signal egress from another constitutes a potential expense for a given BCS operator when an apparent shielding flaw is not, in fact, present in the BCS of that given operator but in another BCS for which the given BCS operator is not responsible. For that reason, signal egress detection equipment acquired by a given BCS operation will be customized and/or calibrated to detect only egress signals from that BCS operator's system. However, the frequencies of both mandatory and optional pilots are readily determinable through use of a spectrum analyzer although that information for an existing OFDM environment is only of interest to a BCS operator in the process of choosing unique frequencies forming a unique pattern or a unique center frequency of fixed PLC-related pattern(s) and to manufacturers, distributors and maintenance service providers of egress signal detection equipment. Therefore, as a general practice, BCS operators do not specify the frequencies or patterns of OFDM pilots even to suppliers of their own egress signal detection equipment and it is the obligation of the egress signal detection equipment suppliers to perform spectral analysis of signals of a given BCS, generally using spectrum analyzers, and to provide egress signal detection equipment that will respond only to signals egressing from a given BCS while BCS operators seek to preserve the ability to discriminate between systems that are geographically proximate to each other.
There are several techniques within present technological capabilities that can be used to detect the scattered pilots which are required by the DOCSIS 3.1 specification. One such method currently preferred will now be described with reference to
Ambient signals are received over antenna 405 and band pas filtered and amplified by a low noise amplifier 410. A tuner 420 and low pass filter 430 are applied to the resulting signal to select a range of frequencies sufficient to encompass several desired scattered pilots. The result is then supplied to and A/D converter 440 and supplied to a DSP or MCU 450.
The processing in the DSP or MCU, again illustrated in a bracket extending therefrom begins with buffering 451. Then a FFT 452 is performed and the magnitude-squared is computed (as before but with different data) to estimate the power spectrum. The logarithm of the power spectrum is then calculated and an inverse FFT (IFFT) calculated and the magnitude-squared again calculated on the result.
This sequence of FFT-log-IFFT calculations (452-454) is generally referred to as the power cepstrum, originally defined in a paper by B. P. Bogert et al. titled “The Quefrency Analysis of Time Series for Echoes: Cepstrum, Pseudo Autocovariance, Cross-Cepstrum and Saphe Cracking” published in Proceedings of the Symposium on Time Series analysis (M. Rosenblatt, Ed.), Chapter 15, pp. 209-243, New York, Wiley, 1963. In the coined terminology, cepstrum is analogous to spectrum, saphe is analogous to phase, quefrency is analogous to frequency, littering is analogous to filtering, etc. in the cepstral domain. The units of quefrency are units of time and thus Cepstral analysis has been used extensively in seismology, radar, modal analysis and analysis of speech and music. The power cepstrum, thus calculated, represents the periodicity in time of a frequency spectrum (or spectral lines thereof) in a manner similar to the way the frequency spectrum represents the periodicity of a time-domain signal. Thus a signal such as DOCSIS 3.1 scattered pilots, which are repetitive at a fixed frequency interval, produces a cepstral line similar to the spectral line produced by a time domain sinusoid and at a point corresponding to the frequency interval at which the scattered pilots occur or recur which, when coincident with the periodicity of scattered plots, is unambiguously indicative of an OFDM signal.
Referring again to
As alluded to above, this technique of detecting scattered pilots discussed above in connection with
In view of the foregoing, it is seen that the invention provides techniques for detecting egressing OFDM signals with improved robustness of detection and without requiring marker signals of any kind or particular behavior which could potentially interfere with data transmission. This is accomplished by detecting specified characteristics of any or any combination of pilots specified (e.g. optional pilots) or required (e.g. PLC-based or scattered pilots) for OFDM transmission and which are thus uniquely characteristic of any OFDM signal that is present in a received ambient signal. Discrimination between overbuilt BCSs can be based on differences in PLC frequencies as reflected in frequencies of a fixed pattern of required pilots, patterns of optional fixed frequency pilots or the frequency band of OFDM blocks between BCSs which can be freely chosen for any OFDM block within the BCS bandwidth.
While the invention has been described in terms of a single preferred embodiment, those skilled in the art will recognize that the invention can be practiced with modification within the spirit and scope of the appended claims.
Claims
1. A method of detecting an orthogonal frequency division multiplex signal among ambient electromagnetic signals comprising steps of
- receiving an ambient electromagnetic signal,
- deriving a spectrum of said ambient electromagnetic signal, and
- analyzing said spectrum to determine one or more characteristics of patterns of locations or intervals of spectral peaks.
2. The method as recited in claim 1, wherein said characteristics of patterns of locations of spectral peaks comprise PHY Link Channel (PLC) related pilots.
3. The method as recited in claim 2, wherein said PLC related pilots are at predetermined frequency intervals from a PLC frequency band.
4. The method as recited in claim 2, wherein said patterns of locations of spectral peaks are:
- ±550 kHz for a narrow PLC or ±750 kHz for a wide PLC from the center frequency of the PLC or 350 kHz from the PLC edges,
- ±1000 kHz for a narrow PLC or ±1200 kHz for a wide PLC from the center frequency of the PLC or 800 kHz from the PLC edges,
- ±1,550 kHz for a narrow PLC or ±1,750 kHz for a wide PLC from the center frequency of the PLC or 1,350 kHz from the PLC edges, and
- ±2,150 kHz for a narrow PLC or ±2,350 kHz for a wide PLC from the center frequency of the PLC or 1950 kHz from the PLC edges.
5. The method as recited in claim 1, wherein said one or more characteristics of patterns of locations or intervals of spectral peaks include
- optional pilot frequencies at selected frequency intervals from each other.
6. The method as recited in claim 5, wherein said optional pilot signals are placed at fixed frequency intervals of 1.6 MHZ and/or 6.4 MHZ or multiples thereof.
7. The method as recited in claim 1, wherein said characteristics of patterns of locations of spectral peaks comprise peaks of varying frequency at frequency separations of a predetermined frequency wherein said varying frequency repeats at a predetermined time interval.
8. The method as recited in claim 7, wherein said predetermined time interval is a multiple of a symbol rate of a broadband communication signal of interest.
9. The method as recited in claim 8, wherein said multiple of said symbol rate is 64 or 128 symbols.
10. The method as recited in claim 7 wherein said predetermined time interval is 2.56 msec.
11. The method as recited in claim 7, wherein said characteristics of patterns of locations of spectral peaks comprise a cepstral line within a cepstral window.
12. The method as recited in claim 1, wherein said spectral peaks are binary phase shift keying modulated.
13. The method as recited in claim 12, wherein said modulation of said spectral peaks is synchronized.
14. The method as recited in claim 11, wherein said spectral peaks are binary phase shift keying modulated by a pseudo-random binary signal sequence.
15. The method as recited in claim 14, wherein said pseudo-random binary signal sequence is derived from a linear shift register with feedback from predetermined stages.
16. The method as recited in claim 1, wherein said spectral peaks are unmodulated carrier wave signals.
17. The method as recited in claim 1, wherein said patterns of locations differ between geographically proximate broadband communications systems.
18. A method of authenticating a detected signal as an egress signal from a broadband communication system (BCS) system wherein a signal carried by said BCS system includes a plurality of pilot or marker signals, said pilot or marker signals being of different types characterized by frequency, pattern of frequency or rate of recurrence, said method comprising a step of
- detecting at least two types of said marker or pilot signals having particular frequencies, patterns of frequencies or rates of recurrence.
19. The method as recited in claim 18 wherein said pilot signals comprise PLC-related pilots, optional pilots and scattered pilots.
20. Apparatus for detecting an egress signal from a broadband communication system, said apparatus comprising
- a tunable receiver to receive ambient electromagnetic signals,
- a converter for digitizing samples of said ambient electromagnetic signals received by said tunable receiver, and
- a processor configured to perform computations of a computation of a magnitude squared fast Fourier transform of said digitized samples, a logarithm of results of said magnitude squared fast Fourier transform, and an inverse fast Fourier transform of results of said computation of said logarithm.
21. Apparatus as recited in claim 20, wherein said processor or another processor included in said apparatus is configured to additionally perform computations of
- a magnitude squared fast Fourier transform of said digitized samples, and
- cross-correlation of results of said magnitude squared fast Fourier transform with a template.
Type: Application
Filed: May 20, 2014
Publication Date: Nov 26, 2015
Applicant: ComSonics, Inc. (Harrisonburg, VA)
Inventor: John Murphy (Mt. Crawford, VA)
Application Number: 14/282,381