Şehiriçi Haberleşme Kablolarının, Yf Parametrelerinin İncelenmesi Ve 2 Mbit/s'de Kullanılabilirliğinin Araştırlması
Şehiriçi Haberleşme Kablolarının, Yf Parametrelerinin İncelenmesi Ve 2 Mbit/s'de Kullanılabilirliğinin Araştırlması
Dosyalar
Tarih
1994
Yazarlar
Yılmaz, Güneş
Süreli Yayın başlığı
Süreli Yayın ISSN
Cilt Başlığı
Yayınevi
Fen Bilimleri Enstitüsü
Institute of Science and Technology
Institute of Science and Technology
Özet
Teknolojinin bugünkü seviyesinde gerekli incelemeler yapıp önlemler alındığında, dağıtım kablolarının bir kısmında, bakır iletkeni! çift üzerinden 2.048 MBit/s sayısal haberleşme, teknik açıdan mümkün, ekonomik açıdan uygun ve gelecek 10-15 yıl içerisinde abonelerin büyük çoğunluğu için yeterli olacağı düşünülmektedir. Dolayısıyla mevcut telefon şebekesinde ses frekans dağıtım kablolarının yüksek frekans parametrelerini inceleyip, ne oranda 2 M Bit Is sayısal haberleşmede kullanılabileceğinin araştırılması faydalı olacaktır. Bu çalışmada deney ve incelemeler iki yönde yapılmıştır. 1. Teorik hesaplamaların, pratik ölçmeler ve deney sonuçlarıyla ne derece bağdaştığının incelenmesi; 1 kHz -10 MHz frekans bandında geçerli olmak üzere teorik hesaplamalarla bulunan kablo birincil ve ikincil parametreleri, yakın uç ve uzak uç diyafoni zayıflaması değerleri, dağıtım şebekesinde kullanılan 0.4 mm, 0.5 mm, 0.6 mm ve 0.9 mm çapında bakır iletkenli PD - AP- A ve PDF tipi kablolarda yapılan incelemeler ve deney sonuçlan ile kıyaslanmış, formüllerin çok az hata ile (< %3 veya < %2 ) geçerli olduğu frekans bandları tesbit edilmiştir. Ayrıca hat zayıflaması ve karakteristik empedansı hesaplama formüllerine frekansla değişen düzeltme faktörü ilavesi teklif edilmiştir. Çok çiftli kablolarda, çiftler arası diyafoni etkisi daha net incelenebilmesi için bozucu etkinin şiddetine göre 5 gruba ayrılmış ve her grup için ampirik formüller teklif edilmiştir. 2. Dağıtım Kablolarının 2 M Bit/s Hızında Sayısal Haberleşmeye Uygunluğunun İncelenmesi Telefon şebekesi olarak döşenen ses frekans dağıtım kabloları 1 kHz - 2 MHz arası kullanıldığında, hat zayıflaması yapısal düzgünsüzlükler ve eklerden (muflardan) gelen yansımalar, komşu çiftler arası diyafoni gibi faktörlerin etkisi, yüksek rekanslarda çok daha şiddetlidir. 2 MBit/s sayısal haberleşmede kullanılan HDB3 kodu (fN= 1.024 kHz) yerine, 2B1Q kodu(fN=512 kHz) kullanıldığında, hattaki zayıflama ~%25-30 az olur, diyafoni zayıflaması ise 3-4dB daha yüksektir. Bunun sonucu olarak2B1Q kodlama sistemi, kullanabilen maksimum kablo uzunluklarının ~ % 45 kadar artmasını sağlar. Bu tezde telefon şebekesine döşenen toplam 8 tip, dolgulu ve dolgusuz kablonun gerçek kanal kapasitesinin frekansla değişimi incelenmiş, 512 kHz ve 1024 kHz' de maksimum kullanılabilen uzunluklar tesbit edilmiştir. Aynı temel demet içinde birden fazla ( 2,3,4 ve 5 ) çiftin 2 MBit/s hızında haberleşme kanalı olarak kullanıldığında maksimum uzunlukların ne oranda azaldığı gösterilmiştir. Çift sayısı 10 ile 1200 arası değişen 0.4 mm, 0.5 mm, 0.6 mm ve 0.9 mm bakır telli kablolarda, çiftlerin % 10, % 20,. % 50 ve % 100'ü iletişim hızı 2 MBit / s sayısal haberleşmede kullanıldığında, işaret gürültü oranı 15 dB'den az olmaması şartıyla maksimum hat uzunlukları tesbit edilmiştir.
The total investment in order to install copper wire cable network around the world is over 400 billion dollars. The share of local cables in total investment of telecommunication systems is approximately 20-30 %. This ratio is around 19.6-30.34% for the Turkish PTT between 1985-1990. However, the existent copper telecommunication cables are only used in voice frequency band. Although, the total investment costs are quite high for these cables, they are unable to meet broad band telecommunication service requirements of subscribers with new technological developments. This type of service will be supplied over fiber optic cables in B-ISDN with a transfer speed of over 140 M Bit/s. In order to realize such a system, a long time (around 15-20 years) and a large investment are necessary. On the other hand, a copper conductor telephone network with a large investment value is existent and it is used inefficiently only in norrow frequrency band. Therefore, the existent copper wire telecommunication cable will be the backbone of telecommunication in subscriber line for a long time (15-20 years) % Fiber optic cables A \ 1980 1985 1990 1995 2000 2005 2010 2015 2020 Figure 1.4 The Amount of Different Type of Cables Used in World's Telephone Network So all PTTs investigate new ways to work on this telephone network system in the world. It is clear that the data transfer over 10 MBit/s and higher speeds in subscriber line is not economically and technically suitable. However, when necessary measures are taken in today's technological level it is possible to achieve 2.048 M Bit/s digital communication over copper conductor pairs in most of the local cables. It is thought that this will be economically suitable and will meet the needs of most of the subscribers within next xiv 10-15 years. At this point, it is worthy to study the possibility of using voice frequency cables in 2.048 MBit/s digital communication by investigating high frequncy paremeters. 1. Verification of theoretical calculation with practical measurements and experimental results Before going to details, some informations about different types of insulated voice - frequency local telephone cables, unit composed of definite numbers of cores, main unit and cable core are given. In therotical analysis, primary parameters(loop resistance of pair, self inductance, effective capacity and perditans ), secondary parameters of cables ( attenuation phase constant, characteristic impeadance, group delay), cross-talk between pairs, reflections caused by imperfections through the line has been examined. In a frequency band between 1kHz - 10 MHz, theoretical calculated values of primary and secondary parameters of cables, Far and Nearend crosstalk attenuation and experimental values has been found to be the same with a very small error percentage ( less than 2 or 3 % ) for the local cables with different conductor diameters (0.4, 0.5, 0.6 and 0.9 mm in PD and PDF type cables). Depending on the changing characteristics of the parameters, the studies have been carried out in three different frequency bands, a) Voice - frequency band ( 300 Hz-10 kHz ) b) Medium frequency band ( 10 kHz-100 kHz ) c) High frequency band ( higher than 100 kHz ). The bounderies of these frequency bands change depending on conductor diameter and type of the cable. For the evaluation of the loop resistance, 1 kHz - 10 MHz band has been divided into 5 parts and different formulas have been applied for each part. Line attenuation end cross-talk calculations include different constant values depending on the cable type and frequency. The line attenuation and phase constant increase continiously by frequency. The change of eti, az, as by frequency shown in figure (2.21) are according to equations (2.55), (2.61), (2.68). It is shown that ag = a,2 for low frequencies and ocg = as for high frequencies. The experiments show that cm js the nearest to real value. The difference between ag ( real line attenuation ) and ot2 is between % 0.5 - % 1 for low frequency. When frequency increase, this difference also increase that occur %8 at 512 kHz and % 10 at 1024 kHz. As required correction factors which change by frequency are added to oi2 values. Empirical formulas for different types of cables are as follows: xv For Unfilled Cables 0.8 0.4/0.7 ag = ct2 +? 2.0. (f) [dB] 0.7 0.5/0.86 ag = a2 +2.3.(f) [dB] 0.6 0.6/1.1 ag = (X2 + 2.5. (f) [dB] 0.55 0.9/1.64 ag = (X2 + 2.5. (f) [dB] For Filled Cables 0.9 0.4/0.84 ag = a2 +2.5.(f) [dB] 0.65 0.5/1.04 ag = (X2 +2.6.(f) [dB] 0.7 0.6/1.4 ag = a2 + 2.8. (f) [dB] 0.8 0.9/2.0 ag = a2 + 2.8. (f) [dB] The frequency must be taken in MHz for correction factor. To study the crosstalk effect in defail between pairs in multipair cables, the crosstalk sources are divided into 5 groups depending on the level of distortion effect. Empirical formulas are also applied for each group. The groups are given below a ) Cross-talk loss between two pairs in a quad, am =49.2- 10.5 Lg(f) [ dB ] atı =44.0 -17.2 Lg(f) [dB] b ) Cross-talk loss between two pairs in different quads of a unit, an2 = 53.0- 17.1 Lg(f) [dB] afe = 52.0- 17.4 Lg(f) [dB] c ) Cross - talk loss between pairs of different units, in main units, an3 = 62.0- 15.5 Lg(f) [dB] a*, = 56.0- 18.4 Lg(f) [dB] d ) Cross-talk loss between pairs of adjacent main units, aT = 78.0- 14.4 Lg(f) [dB] au =67.5- 19.5 Lg(f) [dB] XVI e ) Cross-talk loss between pairs of non-adjacent main units. ans = 82.0- 13.5 Lg(f) [ dB ] afe = 71.0- 21.0 Lg(f) [dB] Where an - near end crosstalk [ dB ] af - far end crosstalk [ dB ] f - [MHz] 2. Study on application properties of local telephone cables in digital communication at a speed of 2 MBit/s. When the voice frequency local cables which are installed as telephone network are used in the range of 1 kHz - 2 MHz, the damaging effects from factors like line attenuation, structural irregularities, reflections from joints and cross-talk between adjacent pairs are more distinctive at high frequencies. The line attenuation decreases by ~25 -30 % and cross-talk loss increases by 3-4dB when the code HDB3 (fN = 1024 kHz) is replaced by the code2B1Q (fN = 512 kHz) digital communication at 2 MBit/s. As a result of this, the coding system 2B1Q allows to increase the maximum allowable cable length by ~ 45 %. In this thesis, the total of 8 types of filled and unfilled cables which have been installed in a telephone network have been examined for their real channel capacity variation with the frequency and maximum allowable lengths have been found at 512 and 1024 kHz. The decreasing ratio of maximum lengths has been shown when more than one pair ( 2, 3, 4 and 5 ) has been used as a telecommunication channel at 2 Mbit / s within the same subunit. Maximum line lengths with the condition that the signal to noise ratio should not be less than 15 dB, have been determined in 0.4, 0.5, 0.6 and 0.9 diameter copper wire multipair cables, where the pair number is between 10 and 1200, and 10%, 20%, 50%, and 100% of pairs are used in 2 Mbit/s. For different types of cables these lengths are given in the following tables. Table 1. The maximum loop lengths can be used according to 2B1Q code system for 2.048 MBit/s transmission XVII Table 2. The maximum loop lengths can be used according to HDB3 code system for 2.048 M Bit/s transmission Table 3. The max. loop length which 2.048M Bit/s can be transmitted % 50 and %100 of pairs for 0.4 and 0.5 mm wire diameter on of cables according to 2B1Q coding system Table 4. The max. length which 2.048 M Bit/s can be transmitted on %50 and %100 of pairs for 0.4 and 0.5 mm wire diameter of cable according to HDB3 coding system The lengths given in table 5 valid for the homogeneous separation of HF pairs in groups. In practical applications there is also another factor which limits the loop length. It is required that the line attenuation between 2 repeaters or subscriber lines must be less than 40 dB. The max.lengths used for different types of cables according to this limitation are given on Table 5. XVI! Table 5. The lengths of which the line attenuation is 40 dB at 512 kHz and 1024 kHz for different types of cables ( L - km ) When using 2.048 MBit/s communication on local telephone network, 50% of the subscriber line lengths can be used with HDB3 coding system, on the other hend 73% of them can be used with 2B1Q coding system.
The total investment in order to install copper wire cable network around the world is over 400 billion dollars. The share of local cables in total investment of telecommunication systems is approximately 20-30 %. This ratio is around 19.6-30.34% for the Turkish PTT between 1985-1990. However, the existent copper telecommunication cables are only used in voice frequency band. Although, the total investment costs are quite high for these cables, they are unable to meet broad band telecommunication service requirements of subscribers with new technological developments. This type of service will be supplied over fiber optic cables in B-ISDN with a transfer speed of over 140 M Bit/s. In order to realize such a system, a long time (around 15-20 years) and a large investment are necessary. On the other hand, a copper conductor telephone network with a large investment value is existent and it is used inefficiently only in norrow frequrency band. Therefore, the existent copper wire telecommunication cable will be the backbone of telecommunication in subscriber line for a long time (15-20 years) % Fiber optic cables A \ 1980 1985 1990 1995 2000 2005 2010 2015 2020 Figure 1.4 The Amount of Different Type of Cables Used in World's Telephone Network So all PTTs investigate new ways to work on this telephone network system in the world. It is clear that the data transfer over 10 MBit/s and higher speeds in subscriber line is not economically and technically suitable. However, when necessary measures are taken in today's technological level it is possible to achieve 2.048 M Bit/s digital communication over copper conductor pairs in most of the local cables. It is thought that this will be economically suitable and will meet the needs of most of the subscribers within next xiv 10-15 years. At this point, it is worthy to study the possibility of using voice frequency cables in 2.048 MBit/s digital communication by investigating high frequncy paremeters. 1. Verification of theoretical calculation with practical measurements and experimental results Before going to details, some informations about different types of insulated voice - frequency local telephone cables, unit composed of definite numbers of cores, main unit and cable core are given. In therotical analysis, primary parameters(loop resistance of pair, self inductance, effective capacity and perditans ), secondary parameters of cables ( attenuation phase constant, characteristic impeadance, group delay), cross-talk between pairs, reflections caused by imperfections through the line has been examined. In a frequency band between 1kHz - 10 MHz, theoretical calculated values of primary and secondary parameters of cables, Far and Nearend crosstalk attenuation and experimental values has been found to be the same with a very small error percentage ( less than 2 or 3 % ) for the local cables with different conductor diameters (0.4, 0.5, 0.6 and 0.9 mm in PD and PDF type cables). Depending on the changing characteristics of the parameters, the studies have been carried out in three different frequency bands, a) Voice - frequency band ( 300 Hz-10 kHz ) b) Medium frequency band ( 10 kHz-100 kHz ) c) High frequency band ( higher than 100 kHz ). The bounderies of these frequency bands change depending on conductor diameter and type of the cable. For the evaluation of the loop resistance, 1 kHz - 10 MHz band has been divided into 5 parts and different formulas have been applied for each part. Line attenuation end cross-talk calculations include different constant values depending on the cable type and frequency. The line attenuation and phase constant increase continiously by frequency. The change of eti, az, as by frequency shown in figure (2.21) are according to equations (2.55), (2.61), (2.68). It is shown that ag = a,2 for low frequencies and ocg = as for high frequencies. The experiments show that cm js the nearest to real value. The difference between ag ( real line attenuation ) and ot2 is between % 0.5 - % 1 for low frequency. When frequency increase, this difference also increase that occur %8 at 512 kHz and % 10 at 1024 kHz. As required correction factors which change by frequency are added to oi2 values. Empirical formulas for different types of cables are as follows: xv For Unfilled Cables 0.8 0.4/0.7 ag = ct2 +? 2.0. (f) [dB] 0.7 0.5/0.86 ag = a2 +2.3.(f) [dB] 0.6 0.6/1.1 ag = (X2 + 2.5. (f) [dB] 0.55 0.9/1.64 ag = (X2 + 2.5. (f) [dB] For Filled Cables 0.9 0.4/0.84 ag = a2 +2.5.(f) [dB] 0.65 0.5/1.04 ag = (X2 +2.6.(f) [dB] 0.7 0.6/1.4 ag = a2 + 2.8. (f) [dB] 0.8 0.9/2.0 ag = a2 + 2.8. (f) [dB] The frequency must be taken in MHz for correction factor. To study the crosstalk effect in defail between pairs in multipair cables, the crosstalk sources are divided into 5 groups depending on the level of distortion effect. Empirical formulas are also applied for each group. The groups are given below a ) Cross-talk loss between two pairs in a quad, am =49.2- 10.5 Lg(f) [ dB ] atı =44.0 -17.2 Lg(f) [dB] b ) Cross-talk loss between two pairs in different quads of a unit, an2 = 53.0- 17.1 Lg(f) [dB] afe = 52.0- 17.4 Lg(f) [dB] c ) Cross - talk loss between pairs of different units, in main units, an3 = 62.0- 15.5 Lg(f) [dB] a*, = 56.0- 18.4 Lg(f) [dB] d ) Cross-talk loss between pairs of adjacent main units, aT = 78.0- 14.4 Lg(f) [dB] au =67.5- 19.5 Lg(f) [dB] XVI e ) Cross-talk loss between pairs of non-adjacent main units. ans = 82.0- 13.5 Lg(f) [ dB ] afe = 71.0- 21.0 Lg(f) [dB] Where an - near end crosstalk [ dB ] af - far end crosstalk [ dB ] f - [MHz] 2. Study on application properties of local telephone cables in digital communication at a speed of 2 MBit/s. When the voice frequency local cables which are installed as telephone network are used in the range of 1 kHz - 2 MHz, the damaging effects from factors like line attenuation, structural irregularities, reflections from joints and cross-talk between adjacent pairs are more distinctive at high frequencies. The line attenuation decreases by ~25 -30 % and cross-talk loss increases by 3-4dB when the code HDB3 (fN = 1024 kHz) is replaced by the code2B1Q (fN = 512 kHz) digital communication at 2 MBit/s. As a result of this, the coding system 2B1Q allows to increase the maximum allowable cable length by ~ 45 %. In this thesis, the total of 8 types of filled and unfilled cables which have been installed in a telephone network have been examined for their real channel capacity variation with the frequency and maximum allowable lengths have been found at 512 and 1024 kHz. The decreasing ratio of maximum lengths has been shown when more than one pair ( 2, 3, 4 and 5 ) has been used as a telecommunication channel at 2 Mbit / s within the same subunit. Maximum line lengths with the condition that the signal to noise ratio should not be less than 15 dB, have been determined in 0.4, 0.5, 0.6 and 0.9 diameter copper wire multipair cables, where the pair number is between 10 and 1200, and 10%, 20%, 50%, and 100% of pairs are used in 2 Mbit/s. For different types of cables these lengths are given in the following tables. Table 1. The maximum loop lengths can be used according to 2B1Q code system for 2.048 MBit/s transmission XVII Table 2. The maximum loop lengths can be used according to HDB3 code system for 2.048 M Bit/s transmission Table 3. The max. loop length which 2.048M Bit/s can be transmitted % 50 and %100 of pairs for 0.4 and 0.5 mm wire diameter on of cables according to 2B1Q coding system Table 4. The max. length which 2.048 M Bit/s can be transmitted on %50 and %100 of pairs for 0.4 and 0.5 mm wire diameter of cable according to HDB3 coding system The lengths given in table 5 valid for the homogeneous separation of HF pairs in groups. In practical applications there is also another factor which limits the loop length. It is required that the line attenuation between 2 repeaters or subscriber lines must be less than 40 dB. The max.lengths used for different types of cables according to this limitation are given on Table 5. XVI! Table 5. The lengths of which the line attenuation is 40 dB at 512 kHz and 1024 kHz for different types of cables ( L - km ) When using 2.048 MBit/s communication on local telephone network, 50% of the subscriber line lengths can be used with HDB3 coding system, on the other hend 73% of them can be used with 2B1Q coding system.
Açıklama
Tez (Doktora) -- İstanbul Teknik Üniversitesi, Fen Bilimleri Enstitüsü, 1994
Thesis (Ph.D.) -- İstanbul Technical University, Institute of Science and Technology, 1994
Thesis (Ph.D.) -- İstanbul Technical University, Institute of Science and Technology, 1994
Anahtar kelimeler
Kablolar,
Parametreler,
Telefon hattı,
İletişim ağları,
Cables,
Parameters,
Telephone line,
Communication networks