Article 41576 of comp.dcom.telecom: Path: mri-gw!psinntp!psinntp!rutgers!sgigate.sgi.com!sgiblab!swrinde!howland.reston.ans.net!spool.mu.edu!uwm.edu!lll-winken.llnl.gov!telecom-request Date: Mon, 19 Dec 1994 10:35:29 -0500 From: gordon@torrie.org (Gordon Torrie) Newsgroups: comp.dcom.telecom Subject: Re: Help Converting V&H Coordinates to Longitude and Latitude Message-ID: Organization: Torrie Communication Services Sender: telecom@eecs.nwu.edu Approved: telecom@eecs.nwu.edu X-Submissions-To: telecom@eecs.nwu.edu X-Administrivia-To: telecom-request@eecs.nwu.edu X-Telecom-Digest: Volume 14, Issue 456, Message 1 of 9 Lines: 243 C. Edward Chow writes: > I am working on a project that needs to convert switching > nodes' locations expressed in terms of V&H coordinates to those > in Longitude and Latitude coordinates. I checked with > telecommunications references in the library but can not find > the definition of telephone network V&H coordinates. Can > someone help pointing to the right references or explain it? Here are four messages which have appeared in comp.dcom.telecom in the past and which describe how to perform the V&H <=> Lat./Long. conversions including source code routines in C and Fortran. From: Thomas B. Libert Subject: Re: H & V Distance Computing Algorithm Wanted Date: Sun, 26 Sep 93 14:43:23 EDT Reply-To: tom@comsol.com Organization: Computing Solutions Inc., Ann Arbor, Michigan X-Telecom-Digest: Volume 13, Issue 666, Message 8 of 16 The H & V numbers are computed from the Donald Elliptical projection, a two-point equidistant projection designed by Jay K. Donald of AT&T around 1956. The projection allows you to use the simple cartesian distance formula to compute a reasonably good approximation for the true distance. Just compute sqrt((h2 - h1)^2 + (v2 - v1)^2), and multiply by a scale factor (ten miles? something like that.) From: StuJeffery@cup.portal.com Subject: Re: H & V Distance Computing Algorithm Wanted Date: Sun, 26 Sep 93 22:33:32 PDT X-Telecom-Digest: Volume 13, Issue 666, Message 9 of 16 Jimmy Gauvin (jimmy@cerberus.ulaval.ca) writes: > Can somebody please tell me how to calculate the distance between two > NPA-NXXs given their H & V coordinates? The following is the answer in two forms: 1. as a text statement: D = Square Root of (((V1 - V2)squared + (H1 - H2)squared)/10) where D is in miles and V and H are values in the V and H Coordinate System. For example the distance from San Diego ( 9462 7632) to San Franciso (8493 8717) is: 460.02, to Sacramento (8303 8581) is: 473.70, to Los Angles (9213 7878) is: 110.69. 2. as a C program: /* * This program computes distance in miles between * two points given their V and H coordinates. * */ #include #include main() { float D; int V1, V2, H1, H2; while ( 1 == 1) { printf( "enter V and H for first point:\n" ); scanf( "%i%i", &V1, &H1 ); printf( "enter V and H for second point:\n" ); scanf( "%i%i", &V2, &H2 ); D = sqrt((pow ((V1-V2), 2) + pow ((H1-H2), 2) ) / 10 ); printf( "Distance in miles = %7.2f \n", D); printf (" \n"); } } ------------- tear here ---------------- Good luck, Stuart Jeffery 415-966-8199 Subject: V & H FORTRAN Routines Date: Jul 01 1990 From: Mike.Riddle@f27.n285.z1.fidonet.org The following routines were mentioned by Jim Riddle in a recent Digest article. The response has already been substantial; however, Jim has no easy way of sending files through the Internet. Since several people have already responded, perhaps you couuld run them in the Digest or mention they are available wherever you put them in the archives? C TO CONVERT EITHER EARTH-CENTERED TO LAT/LON OR VICE VERSA SUBROUTINE M2CONV (OPT,NUMPTS,LATS,LONS,XLAIST,YLIST,ZLIST) INTEGER*4 OPT,NUMPTS,I REAL*8 LATS(NUMPTS),LONS(NUMPTS),THETA,PHI,DTR REAL*8 XLIST(NUMPTS),YLIST(NUMPTS),ZLIST(NUMPTS) PI=3.141592653589793238462643 DTR=PI/180. IF (OPT .EQ. 1) THEN DO FOR I = 1,NUMPTS THETA = LONS(I) * DTR PHI = (90.0D+0 - LATS(I)) * DTR XLIST(I) = DCOS(THETA) * DSIN(PHI) YLIST(I) = DSIN(THETA) * DSIN(PHI) ZLIST(I) = DCOS(PHI) ENDDO ELSEIF (OPT .EQ. 2) THEN DO FOR I = 1,NUMPTS LATS(I) = 90.0D+0 - DACOS(ZLIST(I)) / DTR IF (XLIST(I) .EQ. 0.0D+0) THEN IF (YLIST(I) .LT. 0.0D0+0) THEN LONS(I) = -90.0D+0 ELSE LONS(I) = 90.0D0+0 ENDIF ELSE LONS(I) = DATAN(YLIST(I)/XLIST(I)) / DTR IF (XLIST(I) .LT. 0.0D0+0) THEN IF (YLIST(I) .LT. 0.0D0+0) THEN LONS(I) = LONS(I) - 180.0D0+0 ELSE LONS(I) = LONS(I) + 180.0D0+0 ENDIF ENDIF ENDIF ENDDO ENDIF END SUBROUTINE LLAXYZ(LAT,LON,ALT,Y,Y,Z) C SUBROUTINE TO CALCULATE ECI FROM LAT/LON/ALT C C RE IS RADIUS OF THE EARTH C RE = 3437.5 NAUTICAL MILES MORE OR LESS C 1 NAUTICAL MILE = 1852 METERS (EXACT) C 1 STATUTE MILE = 1609.344 METERS (EXACT) C SUBROUTINE _DOES_ REQUIRE ALTITUDE -- CHECK THE LOCAL AIRPORT (!) C INPUT REQUIRED IS LAT, LON, ALT OF POINT AND RETURNS X, Y, Z C REAL LAT,LON,ALT RE = 3437.75 C RE = 3437.75 * 1.852 FOR KILOMETERS C RE = 3437.75 * 1.852/1609.344 FOR STATUTE MILES C ON MOST MACHINES YOU MAY WANT TO GO DOUBLE PRECISION, BY THE WAY RANGE = RE + ALT CLAT = COS(LAT) SLAT = SIN(LAT) CLON = COS(LON) SLON = SIN(LON) Z = RANGE * SLAT X = RANGE * CLAT * CLON Y = RANGE * CLAT * SLON RETURN END C CONVERTS LAT AND LONG TO EUCLIDEAN COORDINATES SUBROUTINE M2EUCL REAL*4 DEGRAD, PI, ECLX(N), ECLY(N), ECLZ(N) REAL*4 LOND(N), LATD(N) INTEGER*4 NUMPTS,I PI=3.141592653589793238462643 DEGRAD = PI/180. DO FOR I = 1,NUMPTS ECLX(I) = COS(LOND(I)*DEGRAD)*COS(LATD(I)*DEGRAD) ECLY(I) = SIN(LOND(I)*DEGRAD)*COS(LATD(I)*DEGRAD) ECLZ(I) = SIN(LATD(I)*DEGRAD) ENDDO RETURN END SUBROUTINE M2DIST(XPT,YPT,ZPT,NUMLST) INTEGER*4 NUMLST,N,INDEX C CALCULATE DISTANCES (GREAT CIRCLE) FROM A LIST OF POINTS TO C A FOCAL POINT C XPT, YPT, ZPT ARE X, Y, Z COORDINATES OF THE FOCAL POINT C XLST, YLST, ZLST ARE LISTS OF X, Y AND Z COORDINATES C RADIUS IS RADIUS OF THE EARTH REAL*4 XPT,YPT,ZPT,CSQRED,CTHETA,XLST(N),YLST(N),ZLST(N),DLST(N) DO FOR INDEX = 1, NUMLST CSQRED = (XLST(INDEX) - XPT)**2 + (YLST(INDEX) - YPT)**2 + 1 (ZLST(INDEX) - ZPT)**2 CTHETA = 1.0 - CSQRED/2.0 DLST(INDEX) = ACOS(CTHETA) * RADIUS ENDDO RETURN END --- Ybbat (DRBBS) 8.9 v. 3.11 r.3 [1:285/27@fidonet] The Inns of Court 402/593-1192 (1:285/27.0) Subject: Comments For V & H FORTRAN Routines Date: Jul 01 1990 From: Mike.Riddle@f27.n285.z1.fidonet.org [Moderator's Note: This message was intended to accompany the V&H Fortran routines which appeared in Digest # 458 early Wedensday. Unfortunately it was delayed in transmission and arrived here later. PT] The file in the accompanying message contains the routines I mentioned. It contains several Fortran subroutines which should be quasi-obvious to implment. If you don't code Fortran, you can probably understand then from either BASIC or Pascal; if you can't comprehend the math involved, then you may be in over your head already. When using the coordinate system, it is NOT necessary that you have the same X-Y grid as the phone company so long as you are using statute miles, nautical miles, furlongs, picometers or whatever consistent with their distancing measurement. The most convienent way to do it is probably to use real-Earth latitudes and longtitudes and PAY ATTENTION TO WHICH ROUTINES USE DEGREES AND WHICH ARE IN RADIANS. If you really want a polished, finished product, I really can't comply as I am under some conflict of interest restrictions (sounds nebulous because it is). Thanks for the bevy of responses I received. Jim R. --- Ybbat (DRBBS) 8.9 v. 3.11 r.3 [1:285/27@fidonet] The Inns of Court 402/593-1192 (1:285/27.0) --- Through FidoNet gateway node 1:16/390 Mike.Riddle@f27.n285.z1.fidonet.org ---------------- gordon@torrie.org Gord Torrie