Communication network intialization apparatus and method for fast GPS-based positioning

BACKGROUND OF THE INVENTION

(a) Field of the Invention

The present invention relates generally to a method and apparatus for initializing a Global Positioning System (GPS) receiver for use in a communication network, and more particularly, to a method and apparatus for initializing a GPS receiver forming part of a communication network in such a manner that facilitates rapid determination of the position of the GPS receiver upon startup of the GPS receiver.

(b) Description of Related Art

Typically, a GPS receiver relies on GPS satellite trajectory parameters stored in memory during recent operation, a time estimate from a running clock or user input, and a position estimate from memory or user input to perform a startup faster than a “cold start”. If any of this information is missing, a cold start will be necessary and the time to first fix (TTFF) may be 1-3 minutes.

In general, one common strategy for a GPS receiver integrated into a communication network is to either continually track GPS satellites, or cycle on at timed intervals to obtain a fix, resynchronize a local clock, and download GPS parameters. The disadvantages to this approach are: 1) longer acquisition times in general; 2) larger power consumption and processing drain; and 3) the need for GPS acquisition at times when terminal is not otherwise in use (in the case of a handheld terminal, this is particularly problematic; when the terminal is not in use it may be stowed somewhere, such as a pocket or briefcase, where GPS satellite visibility is very poor).

Another strategy involves a wholly integrated approach, where a terminal reports intermediate GPS measurements to the network, and the network performs the actual positioning computations. The disadvantages of this approach are 1) increased data transfer from the terminal to the network is needed; 2) complex network computing facilities are required to handle each terminal separately; and 3) the terminal is unable to perform GPS positioning when network is not available.

There are technologies emerging in communication networks, primarily for Emergency-911 systems, such as those developed by Navsys, Inc. and Snaptrack, Inc., that use wholly integrated approaches to determination of terminal positions, meaning that the position determination requires substantial handshaking and cooperation between remote terminals and the network infrastructure. (As used herein, the term “terminal” describes a mobile unit within a communication network, such as a cellular phone.)

SUMMARY OF THE INVENTION

The invention is directed to a comprehensive method of utilizing a communications network (in this particular design, it is a GEM (Geo-Mobile) satellite communication network system) to initialize a Global Positioning System (GPS) receiver (such as one contained within a handheld user terminal) to enable the GPS receiver to quickly acquire GPS satellite signals and perform positioning calculations. By systematically feeding the GPS receiver key pieces of information such as, for example, (GPS satellite trajectories, time estimate, position estimate, and additional positional references), the time to first fix (TTFF) of the GPS receiver can be significantly reduced. One novel aspect of the invention is the series of actions taken by the GEM system to determine and provide, to the GPS receiver, information that otherwise may not be available.

The present invention is useful for a Geo-mobile (GEM) satellite phone network, where any and all terminals may be initialized at any time by the generic broadcast of the GEM satellite, but could be applied to a broad class of communication networks. A network may require that a terminal (e.g., a wireless telephone) determine and report its position before each call it places with the network. For this reason, fast GPS-based positioning capability within each terminal is required. This invention provides the capability to automatically and simultaneously initialize all terminals with the information required to obtain minimal GPS acquisition times.

It has been determined that minimal position determination times are achievable if the following five pieces of information are available to a terminal, and hence to its integrated GPS receiver:

1. Satellite parameters describing orbital trajectories for all GPS satellites in view to the terminal, used by the terminal to compute satellite positions at a given instant in time (a critical step in the receiver's computation of its own position);

2. GPS time estimate to within a few milliseconds;

3. A rough position estimate to within a few hundred kilometers;

4. A GPS almanac; and

5. Additional positional references.

As will be appreciated by those skilled in the art, providing orbital trajectories for all GPS satellites visible to the terminal is optimal. However, the invention contemplates utilizing orbital trajectories for fewer than all visible GPS satellites as well.

The system in accordance with the invention continuously determines and provides all five of the above pieces of information to all terminals in the coverage area, so that any terminal, at any time, is able to perform a fast position determination.

The GEM system, for example, is designed to provide these key pieces of information as follows:

1. GPS satellite trajectories; Each ground station supports a continually active GPS receiver that tracks all visible GPS satellites, and stores satellite parameters precisely describing their orbits, for all of them. The ground station predicts which GPS satellites are in view to each distinct service area (spot beam), and the GEM satellite broadcasts local trajectory coefficients (computed at the ground station) for those GPS satellites through each spot beam.

2. GPS time: The ground station is synchronized to GPS time, via the ground station's active GPS receiver. GPS time, adjusted for propagation delay to within a few milliseconds, is broadcast with the satellite information.

3. Rough position estimate: The terminal measures relative signal strengths of broadcast channels in neighboring spot beams. Based on these measurements, a position estimate typically accurate to about 100 km can be computed, although a position estimate accuracy up to about 1,000 km may be acceptable.

4. A GPS almanac is downloaded at the ground station and re-broadcast to the coverage area.

5. With further network initialization, GPS acquisition may be possible in poor GPS satellite visibility situations. An extremely accurate time reference (e.g., accurate to within two microseconds) could reduce the number of required GPS satellite signal acquisitions from four to three, provided some degradation in fix accuracy is acceptable. If the communication network can provide precise pieces of position information (such as altitude, and/or distance from a known reference), the number of GPS satellites needed may be further reduced.

In accordance with one aspect of the present invention, a method is provided for initializing a GPS receiver to rapidly acquire GPS satellite signals for establishing a precise estimate of the position of the GPS receiver. The method comprises the steps of: broadcasting a signal representative of orbital trajectories of one or more GPS satellites within view of the GPS receiver; broadcasting a time synchronization signal; calculating a rough estimate of the position of the GPS receiver may be calculated; and inserting the signal representative of orbital trajectories, the time synchronization signal, and the rough estimate of the position of the GPS receiver into the GPS receiver.

Preferably, the time synchronization signal is accurate to within about five milliseconds and the rough estimate of the position of the GPS receiver is accurate to within about 1,000 kilometers.

Preferably, the orbital trajectories signal is broadcast via a satellite, using an idle communication channel. Also preferably, the time synchronization signal is broadcast via a satellite, using an idle communication channel.

Obviously, it is desirable to have a time synchronization signal that is extremely accurate (e.g., to within a few microseconds). However, such accuracy levels are generally not feasible.

In accordance with another aspect of the present invention, a communication network comprises at least one communication station, a plurality of GPS satellites, and a terminal. The communication station includes capability for sending communication signals, GPS satellite trajectory signals, and time synchronization signals to the terminal. The terminal includes apparatus for receiving communication signals from the communication station, apparatus for sending communication signals to the communication station, apparatus for receiving GPS signals from the GPS satellites, and apparatus for processing the GPS satellite trajectory signals and time synchronization signals to rapidly determine the position of the terminal.

In accordance with yet another aspect of the invention, a method of initializing a remote GPS receiver to rapidly acquire GPS satellite signals for establishing a precise estimate of the position of the remote GPS receiver is provided. The method comprises the steps of: providing an active GPS receiver at a fixed location which computes precise GPS satellite trajectories; broadcasting a signal representative of the precise GPS satellite trajectories of one or more GPS satellites within view of the remote GPS receiver; broadcasting a time synchronization signal; calculating a rough estimate of the position of the remote GPS receiver; and inserting the signal representative of orbital trajectories, the time synchronization signal, and the rough estimate of the position of the remote GPS receiver into the remote GPS receiver.

In accordance with still another aspect of the invention, a terminal for use in a communication network is provided. The communication network comprises at least one communication station and a plurality of GPS satellites. The communication station includes apparatus for sending communication signals, GPS satellite trajectory signals, and time synchronization signals to the terminal. The terminal comprises an apparatus for receiving communication signals from the communication station, an apparatus for sending communication signals to the communication station, an apparatus for receiving GPS signals from the GPS satellites, and an apparatus for processing the GPS satellite trajectory signals and time synchronization signals to rapidly determine the position of the terminal.

In accordance with yet another aspect of the invention, a gateway station for use in a communication network is provided. The communication network comprises at least one terminal, a communications satellite and a plurality of GPS satellites. The gateway station comprises a continually-tracking GPS receiver, apparatus for computing GPS satellite trajectory data for GPS satellites visible to each terminal, apparatus for computing a rough position estimate for each terminal, and apparatus for sending signals representing GPS satellite trajectory data, time synchronization signals, and signals representing rough position estimates to each terminal.

The invention disclosed herein has the distinct advantage that the terminals possess self-contained fully functional GPS receivers. The invention acts as a supplement to greatly accelerate GPS receiver operation, not replace it. Thus, each terminal has full GPS functionality even when the communication network is not available. In addition, the handshaking required to compute a terminal position is reduced to a 1-way generic broadcast from the communication network to the terminals. All terminals in a given geographic cluster of up to a few hundred km receive the same information and proceed with the accelerated GPS acquisitions. No information is required to be transmitted from the terminal to the network. The position, when computed, is available for immediate display to the user of the terminal, and may be transmitted to the network if desired.

By using the invention, the GEM system will automatically provide the GPS receiver with enough information to optimize startup, regardless of what is in memory, without a running clock, and with no user intervention. The result is a more consistent, and much faster initial position fix, that typically takes only a few seconds.

The invention provides enhanced customer satisfaction and power savings in the terminal, which is especially important where the terminal is a portable handset having a finite battery charge. It is, further, particularly effective whenever a user tries to place a phone call in a network that requires a GPS position fix before a call is allowed. GPS positioning functionality consistently delaying call setup for 1-3 minutes would have a devastating effect on product quality. With use of the present invention, the GPS process should delay registration or call setup no more than a few seconds. The second benefit of the invention is that once a GPS position is obtained, the GPS receiver may be turned off. A GPS receiver may consume up to about 0.5 W while on, and use of the invention would greatly reduce the amount of on-time, and therefore power consumption.

The invention itself, together with further objects and attendant advantages, will be best understood by reference to the following detailed description, taken in conjunction with the accompanying drawing.

BRIEF DESCRIPTION OF THE DRAWING

FIG. 1

is a schematic representation of a satellite communication network in accordance with the present invention;

FIG. 2

is a schematic representation of an illumination pattern of overlapping spot beams projected on the Earth by a communication satellite in accordance with the present invention;

FIG. 3

is a schematic representation of the apparent elevation of a GPS satellite forming part of the satellite communication network of

FIG. 1

;

FIG. 4

is a f low diagram illustrating a method, in accordance with the present invention, of initializing a GPS receiver for a rapid position determination;

FIG. 5

is a schematic depiction of a satellite and spot beams emanating from the satellite, showing the increasing diameter of the projection of each spot beam on the Earth, as a function of latitude;

FIG. 6

is a schematic depiction of a satellite and spot beams emanating from the satellite, showing a

2

-D plane upon which the spot beams are mapped for a coordinate transformation procedure for use in a first method of determining a user terminal position estimate based on relative power measurements;

FIG. 7

is a schematic depiction of a satellite and a spot beam emanating from the satellite, showing distances and angles for use in the first method of determining the user terminal position estimate based on relative power measurements;

FIG. 8

is a schematic depiction of three neighboring spot beams, depicting a search procedure, by which lines of equal relative power are used to estimate terminal location using the first method of determining the user terminal position estimate based on relative power measurements;

FIG. 9

is a schematic depiction of three neighboring spot beams, showing distances and angles used for the first method of determining the user terminal position estimate based on relative power measurements;

FIG. 10

is a schematic depiction of a second method for estimating the user terminal position based on relative power measurements by finding a set of points of constant relative power equal to the relative power measured from three strongest neighboring spot beams;

FIG. 11

is a schematic depiction of a cluster of seven spot beams on the surface of the earth; and

FIG. 12

is a schematic depiction of the surface of the earth in relation to a two-dimensional plane onto which the spot beams shown in

FIGS. 10 and 11

are translated in accordance with an embodiment of the present invention.

DESCRIPTION OF THE PREFERRED EMBODIMENT

The invention will now be described in connection with a current application of the inventive method to a GEM satellite communication network system. However, the invention is applicable to any wireless communication network.

For each aspect of this invention described below, the general method that could apply to a wide class of communication networks will be described, as well as the specific method designed for use in the GEM system.

As shown in

FIG. 1

, a satellite network

10

in accordance with the present invention includes a Geo-Mobile (GEM) satellite

12

, a plurality of GPS satellites

14

, a gateway

16

, and a plurality of terminals

18

. Each terminal

18

includes a GPS receiver

19

.

Upon powering up (block

54

of FIG.

4

), each terminal

18

immediately receives position data for visible GPS satellites

14

, a time signal, and performs a rough position estimate of the GPS receiver

19

, as shown at block

56

of FIG.

4

. The rough position estimate could be, for example, based simply on knowledge of which spot beam

22

the GPS receiver

19

is located within. (In a terrestrial cellular application, the cells are typically only a few miles across and accordingly, the rough position estimate could be based on knowledge of which cell contains the GPS receiver

19

.)

The GPS receiver

19

in the terminal

18

is initialized and allowed to perform a precise position computation (block

58

of FIG.

4

). The terminal

18

receives the computed position (block

59

of

FIG. 4

) and transmits the computed position to the GEM satellite

12

(block

60

of FIG.

4

).

Providing GPS Satellite Trajectories

General Case

The network

10

maintains at least one continually-tracking GPS receiver

20

, located at the gateway

16

, that downloads and stores GPS satellite parameters, such as ephemeris, ionosphere correction, clock correction, etc., as defined in Navstar's ICD-GPS-200, from each GPS satellite

14

that is in view, as indicated at block

40

in FIG.

4

. GPS satellite orbits are about 12 hours. While a GPS satellite

14

is out of view, last available parameters are stored and used for their validity duration, as indicated at blocks

42

and

44

of FIG.

4

. Or, extrapolation may be used, or multiple continually-tracking GPS receivers

20

may be spread over the entire coverage area and may be networked to accommodate brief periods between the expiration of a GPS satellite's stored data validity period and the return to view of that GPS satellite.

Each gateway

16

maintains valid parameters for as many as possible GPS satellites

14

visible to area of responsibility of the gateway

16

at all times. When a terminal

18

needs to make a position fix, it receives a signal broadcast by the network

10

that communicates which GPS satellites

14

should be in view to the terminal

18

, based on a rough position estimate, and sends the terminal

18

valid parameters (either the parameters stored, or localized trajectory coefficients derived from the stored parameters) for those GPS satellites

14

. In this manner, the terminal

18

receives the necessary GPS information much more quickly than it could receive and process the same information directly from the GPS satellites

14

, as happens in normal GPS receiver operation.

GEM System

The GEM satellite

12

illuminates the coverage area on the surface of the Earth

21

with overlapping spot beams

22

, about 200 km in radius, as shown in FIG.

2

. Satellite trajectory coefficients, computed from the GPS satellite parameters downloaded and stored at each gateway

16

, which enable a terminal GPS receiver to accurately compute GPS satellite position at any instant in the next few minutes, must be sent separately through each spot beam

22

, for the GPS satellites

14

in view to receivers

18

within that spot beam.

As shown in

FIG. 1

, the continually-tracking GPS receiver

20

is maintained at a each ground station (e.g., gateway

16

). From all GPS satellites

14

in view to the gateway GPS receiver

20

at a given time, parameters such as ephemeris, ionosphere correction, clock correction, etc. as defined in Navstar's ICD-GPS-200, are downloaded and stored at every opportunity. When a GPS satellite

14

leaves view of the central gateway

16

, last available parameters are stored and used, until the GPS satellite data becomes too old or the GPS satellite

14

returns to view (within a few hours).

From these parameters, and as indicated at block

46

of

FIG. 4

, GPS satellite positions are computed every second, as described in Navstar's ICD-GPS-200 and reproduced in Table 1, Table 2, and Table 3. A predictive third order curve fit of the type shown in the equations below is computed for each satellite trajectory, based on Lagrange interpolation of several computed positions, spread over a 10 minute interval, starting with the current position at time t

0

, and predicting future positions.

\begin{matrix} {\begin{matrix} {\tilde{x}}_{k} (t) \approx a_{0} + a_{1} (t - t_{0}) + {a_{2} (t - t_{0})}^{2} + {a_{3} (t - t_{0})}^{3} \\ {\tilde{y}}_{k} (t) \approx b_{0} + b_{1} (t - t_{0}) + {b_{2} (t - t_{0})}^{2} + {b_{3} (t - t_{0})}^{3} \\ {\tilde{z}}_{k} (t) \approx c_{0} + c_{1} (t - t_{0}) + {c_{2} (t - t_{0})}^{2} + {c_{3} (t - t_{0})}^{3} \end{matrix}} & Equation 1 \end{matrix}

For up to twelve GPS satellites

14

in possible view to any portion of a spot beam

22

, the trajectory coefficients a

0-3

, b

0-3

, c

0-3

, and t

0

are broadcast repeatedly to that spot beam

22

.

To determine which GPS satellites

14

are in view to each spot beam

22

, the gateway

16

computes a GPS satellite visibility list for each spot beam

22

, based on the spot beam center coordinates (x

r

,y

r

,z

r

) and each GPS satellite

14

position (x

k

,y

k

,z

k

), every five minutes, as indicated at block

48

of FIG.

4

. The computed position data for visible GPS satellites

14

for each spot beam

22

is processed to form trajectory data, and is then sent to the GEM satellite

12

, as indicated at block

50

of FIG.

4

. The GEM satellite

12

continuously transmits, via each spot beam

22

, computed trajectory data for visible GPS satellites

14

, as indicated in block

52

of FIG.

4

.

Any GPS satellite

14

with an elevation angle from a given center of a spot beam

22

at the surface of the Earth

21

that is within the antenna elevation mask of the terminal

18

, is eligible for the visibility list of that spot beam

22

. The elevation of a GPS satellite

14

at (x

k

,y

k

,z

k

), from a reference point (x

r

,y

r

,z

r

) on the surface of the Earth

21

(in this case a spot beam center) is illustrated in

FIG. 3

, and computed as shown in Table 4. The trajectory coefficients a

0-3

, b

0-3

, and c

0-3

for the GPS satellites

14

determined to be visible within a given spot beam

22

are repeatedly broadcast within the spot beam

22

, by the GEM satellite

12

, to initialize any terminal

18

within that spot beam

22

.

TABLE 1

GPS Satellite Ephemeris Parameters

M

0

Mean Anomaly at Reference Time

Δn

Mean Motion Difference from Computed Value

E

Eccentricity

A

1/2

Square Root of the Semi-Major Axis

Ω

0

Longitude of Ascending Node of Orbit Plane at

Weekly Epoch

i

0

Inclination Angle at Reference Time

ω

Argument of Perigee

Ω′

Rate of Right Ascension

i′

Rate of Inclination Angle

C

uc

Amplitude of the Cosine Harmonic Correction

Term to the Argument of Latitude

C

us

Amplitude of the Sine Harmonic Correction Term

to the Argument of Latitude

C

rc

Amplitude of the Cosine Harmonic Correction

Term to the Orbit Radius

C

rs

Amplitude of the Sine Harmonic Correction Term

to the Orbit Radius

C

ic

Amplitude of the Cosine Harmonic Correction

Term to the Angle of Inclination

C

is

Amplitude of the Sine Harmonic Correction Term

to the Angle of Inclination

t

oe

Reference Time Ephemeris

IODE

Issue of Data (Ephemeris)

TABLE 2

GPS Satellite Position Computation, Part I

μ = 3.986 × 10

14

m

3

/sec

2

WGS 84 value of the Earth's

universal gravitational

parameter

Ω′

e

= 7.292 rad/sec

WGS 84 value of the Earth's

rotation rate

A = {(\sqrt{A})}^{2}

Semi-major axis

n_{o} = \sqrt{\frac{μ}{A^{3}}}

Computed mean motion - rad/sec

t

k

= t-t

oe

Time from ephemeris

reference epoch

n = n

0

+ Δn

Corrected mean motion

M

k

= M

0

+ nt

k

Mean anomaly

M

k

= E

k

− esinE

k

Kepler's equation for

eccentric anomaly (solved by

iteration) - radians

ν_{k} = \tan^{- 1} {\frac{\sin ν_{k}}{\cos ν_{k}}} = \tan^{- 1} {\frac{\sqrt{1 - e^{2}} \sin E_{k} / 1 - e \cos E_{k}}{(\cos E_{k} - e) / (1 - e \cos E_{k})}}

True anomaly

E_{k} = \cos^{- 1} {\frac{e + \cos ν_{k}}{1 + e \cos ν_{k}}}

Eccentric anomaly

Φ

k

= ν

k

+ ω

Argument of latitude

TABLE 3

GPS Satellite Position Computation, Part II

δμ

k

= C

μs

sin 2Φ

k

+ C

μc

cos 2Φ

k

Argument of latitude correction

δr

k

= C

rc

cos 2Φ

k

+ C

rs

sin 2Φ

k

Radius correction

δi

k

= C

ic

cos 2Φ

k

+ C

is

sin 2Φ

k

Correction to inclination

μ

k

= Φ

k

+ δμ

k

Corrected argument of latitude

r

k

= A(1 − e cosE

k

) + δr

k

Corrected radius

i

k

= i

o

+ δi

k

+ i't

k

Corrected inclination

x′

k

= r

k

cos μ

k

Position in orbital plane

y′

k

= r

k

sin μ

k

Position in orbital plane

Ω

k

= Ω′ − Ω′

e

)t

k

− Ω′

e

t

oe

Corrected longitude of ascending

node

x

k

= x

k

′ cosΩ

k − y

k

′ cosi

k

sinΩ

k

Earth-fixed coordinate

y

k

= x

k

′ sin Ω

k

− y

k

′ cos i

k

cos Ω

k

Earth-fixed coordinate

z

k

= y

k

′ sin i

k

Earth-fixed coordinate

TABLE 4

Computation of Satellite Elevation as Observed

from Point on Earth's surface

θ = \tan^{- 1} (\frac{y_{r}}{x_{r}})

Angles relating reference point on Earth's surface

to origin (center of Earth)

φ = 90 ° - \tan^{- 1} (\frac{z_{r}}{\sqrt{x_{r}^{2} + y_{r}^{2}}})

{circumflex over (x)}

k

= x

k

− x

r

Translate origin of

ŷ

k

= y

k

− y

r

coordinate system to

{circumflex over (z)}

k

= z

k

− z

r

reference point on

Earth's surface

{tilde over (x)}

k

= {circumflex over (x)}

k

sinθ − ŷ

k

cosθ

Rotation of coordinate

{tilde over (y)}

k

= {circumflex over (x)}

k

cosθ cosφ + ŷ

k

sinθ cosφ − {circumflex over (z)}

k

sinφ

system so that z-axis

{tilde over (z)}

k

= {circumflex over (x)}

k

cosθ sinφ + ŷ

k

sinθ sinφ + {circumflex over (z)}

k

cosφ

points outward from the

Eath at reference point

on surface

EL = \tan^{- 1} (\frac{{\tilde{z}}_{k}}{\sqrt{{\tilde{x}}_{k}^{2} + {\tilde{y}}_{k}^{2}}})

Elevation angle of satellite as seen from reference point on Earth's surface

Performing the calculations set forth in Tables 2-4 enables the network

10

to choose which GPS satellites

14

to broadcast trajectories for (as well as compute the trajectories themselves) for each spot beam

22

. Any terminal

18

within any spot beam

22

may receive all this information for GPS satellites

14

within view of the terminal

18

. The information must be broadcast repeatedly and quickly.

The terminal

18

receives this information from the GEM satellite

12

, and translates the information to conform to inputs accepted by the GPS receiver.

The GPS receiver

19

computes its precise position based upon the received satellite trajectories and the signals received from a plurality of GPS satellites. These GPS satellite signals must be compensated for timing errors (clock offset, relativity, and group delay) as described in NAVSTAR'S ICD-GPS-200.

For each received GPS satellite signal, a “code phase offset” at time t must be computed:

Δ

t

sv

=a

f

0

+a

f

1

(

t−T

0

c

)+

a

f

2

(

t−T

0

c

)

2

+Δt

r

−T

gd

where a

f

0

, a

f

1

, a

f

2

, Δt

r

and T

gd

are all available from the GPS broadcast as described in NAVSTAR'S ICD-GPS-200, and are referenced to time T

oc

, which is also broadcast.

In accordance with one aspect of the present invention, the satellite trajectory broadcast for each GPS satellite is supplemented with these parameters to enable the GPS receiver

19

to compute its position without downloading these parameters directly from the GPS satellites. Broadcast with each satellite's trajectories are (for the same satellite):

Δt

sv

computed for t

o

, the same instant for which the trajectories were computed.

a

f1

with reduced precision, due to the fact that the broadcast signal gives Δ

sv

a time-varying value. Then the GPS receiver

19

may correct its received Δt

sv

for the time elapsed since the original computation of Δt

sv

at time t

0

a

f

2

also causes time variance, but its effects are neglected and it is not broadcast.

Then the GPS receiver computes a corrected code phase offset Δt

sv

(true)=Δt

sv

(received)+a

f

1

(t−t

0

)

Where t=current time

t

0

=time at which Δt

sv

(received) was computed (same time for which trajectory was computed)

Thus, for each satellite in the visibility list, the trajectory information broadcast is shown in Table 5.

TABLE 5

Field

Description

Bits

T

0

Time at which trajectory and code

40

phase offset were computed. In

GPS time of week, in units of 2

−20

sec

Curve fit

0 if t

0

has not changed since last

1

cutover

broadcast; 1 if t

0

has changed

since last broadcast

ID

Satellite ID (1 to 32)

6

Δt

sv

Code phase offset at time t

0

, in

22

units of 2

−28

sec

a

f1

Clock correction received from

11

satellite broadcast. Rounded to

11 bits. Units of 2

−28

sec/sec

a

0

Trajectory parameter from Eq. [1],

24

in units of 2

2

meters

b

0

Trajectory parameter from Eq. [1],

24

in units of 2

2

meters

c

0

Trajectory parameter from Eq. [1],

24

in units of 2

2

meters

a

1

Trajectory parameter from Eq. [1],

18

in units of 2

−5

m/sec

b

1

Trajectory parameter from Eq. [i],

18

in units of 2

−5

m/sec

c

1

Trajectory parameter from Eq. [1],

18

in units of 2

−5

m/sec

a

2

Trajectory parameter from Eq. [1],

13

in units of 2

−12

m/sec

2

b

2

Trajectory parameter from Eq. [1],

13

in units of 2

−12

m/sec

2

c

2

Trajectory parameter from Eq. [1],

13

in units of 2

−12

m/sec

2

a

3

Trajectory parameter from Eq. [1],

8

in units of 2

−19

m/sec

3

b

3

Trajectory parameter from Eq. [1],

8

in units of 2

−19

m/sec

3

c

3

Trajectory parameter from Eq. [1],

8

in units of 2

−19

m/sec

3

TOTAL

269

Providing Time

General Case

The continually-tracking GPS receiver

20

is synchronized to GPS time. The network

10

synchronizes each terminal

18

to GPS time, and the terminal's GPS receiver

19

can be initialized with accurate GPS time. The accuracy of the GPS receiver's time estimate may be controlled by controlling the timing uncertainty of the synchronization between the terminal

18

and the network

10

.

GEM System

GPS time is available from the GPS receivers

20

of the gateways

16

. Each gateway

16

repeatedly broadcasts GPS time of week, referenced to the broadcast frame structure itself, via the GEM satellite

12

, to each spot beam

22

. For each spot beam

22

, the arrival time of the referenced frame edge at the surface of the Earth

21

is estimated based upon the propagation delay through the GEM satellite

12

and the travel time from the satellite to the surface of the Earth

21

. The delay varies across the beam, and the broadcast is designed so the maximum error is minimized (i.e., the average of the maximal and minimal delays is assumed). The time message broadcast to each spot beam

22

is shown in Table

6

. The terminal

18

converts the time values into GPS time (the alternate format is broadcast for bandwidth savings), and relays it to the terminal GPS receiver

19

.

TABLE 6

GPS Time Synchronization Message

Field

Description

Bits

Time Stamp

Minimax time, in GPS time of

40

week, in units of 2

−20

sec.

Frame Number

The number of the GEM frame

19

to which time stamp is

referenced

Total

59

Providing Position Estimate

General Case

The network

10

has some indication of the location of a terminal

18

. For example, in a cellular network, it is known which cell the terminal

18

is in. For a better estimate, the terminal

18

may store its last known GPS location, and the time of that location fix. The terminal

18

makes an intelligent guess at its location based on: the network's estimate, the predicted accuracy of the network's estimate, the last known position, the time elapsed since that position fix, the predicted likelihood of movement since its last position fix, and the statistical expectation of location of the terminal

18

.

GEM System

The terminal

18

estimates its position using the relative power measurements of signals in the three or four strongest available spot beams

22

. The user terminal first gets the locations of the GEM satellite

12

and the centers of each of a group of seven to ten nearest spot beams

22

from the broadcast system information carried in the received GEM signals. With this information and the measured signals' relative strengths from the three or four strongest spot beams, the terminal

18

can then calculate its approximate location. The exact distance from the terminal

18

to the GEM satellite

12

can be determined at the gateway by monitoring the 2-way signal propagation time. This distance can then be used to increase the user terminal position accuracy.

Either of the two methods of position determination estimation set forth in detail below can be used for performing position determination estimation promptly based on the signal relative power measurements of the three or four strongest spot beams

22

.

POSITION DETERMINATION METHODS BASED ON RELATIVE POWER MEASUREMENTS

Each spot beam

22

in a GEM system has the shape of a cone emitted from the GEM satellite

12

. The projection of this cone on the surface of the Earth

21

forms the area served by the spot beam

22

. From the satellite perspective, all spot beams

22

are about 0.695° in diameter, i.e., about 0.695° apart if the GEM satellite

12

was in a zero-inclination orbit. Because the GEM satellite

12

moves from about 6° and −6° latitude through the day due to inclined orbit operation, the beamwidth from the satellite perspective will vary to maintain a constant beam “footprint” on the ground. Because of the curvature of the Earth

21

, spot beams on the ground have diameters that increase as a function of distance from the subsatellite point. Spot beam diameters can vary between 450 km to more than 1200 km at the far edge of the coverage on the Earth

21

. This is shown in FIG.

5

.

The spot beams

22

on the Earth

21

, projected on a plane perpendicular to the satellite radius are all approximately equivalent, independent of inclined orbit operation, as shown in FIG.

2

.

In the GEM system, the GEM satellite

21

is in a nominal geostationary orbit and consequently appears almost stationary relative to the Earth

21

as compared to a low earth-orbit system. In fact, in the GEM system, the GEM satellite

12

moves between about −6 and 6° latitude through the day due to inclined orbit operation. The GEM satellite

12

is located at approximately 35787 km from the surface of the earth

21

. At that altitude, it is permissible to assume a spherical earth with a radius of 6378 km and ignore the altitude of the terminal

18

on the Earth

21

. Two possible methods are proposed for user terminal position determination.

POSITION ESTIMATE METHOD #1

This position estimate algorithm uses the relative BCCH (Broadcast Control Channel) power measurements of three spot beams

22

at a time. It calculates approximate locations for one to twelve combinations of three spot beams

22

depending on the number of neighbors in the cluster. It then averages these approximate locations to get the final averaged approximate location. It is the preferred method because the algorithm estimates the terminal position much more rapidly.

To ease the position acquisition problem, two transformations are first performed to map the three spot beams

22

on earth

21

on a 2-D plane

100

perpendicular to the satellite radius (x′ axis) and centered around a subsatellite point

102

as shown in FIG.

6

. On this new 2-D plane

100

, the three spot beams

22

have approximately equivalent dimensions. The first transformation rotates the axis (x,y,z) by θ

sat

and ψ

sat

such that the new axis (x′,y′,z′) are aligned with the satellite radius that passes through the subsatellite point

102

. If the GEM satellite

12

has a zero-inclination orbit, θ

sat

is the angle between the x axis and the satellite radius x′ passing through the subsatellite point

102

, and ψ

sat

is equal to 0. From then on, only the coordinates of the spot beams

22

of interest are moved around while the axis remains fixed at (x′,y′,z′). The second transformation projects the three spot beams

22

on earth

21

to the 2-D plane

100

centered around the subsatellite point

102

.

To further simplify the problem, two additional transformations are performed. The cluster of three spot beams

22

is first translated on the 2-D plane

100

such that the selected spot beam is located at the center of the plane. The plane is then translated such that the center of the 2-D plane

100

corresponding to the subsatellite point

102

matches the center of the earth

21

. With these four transformations, the number of unknowns is reduced by one: y and z being the unknowns and x being now equal to 0 for all three spot beams

22

. The following algorithm describes the four steps required to map the three spot beams

22

on this 2-D plane. As a general rule for the equations and figures through the rest of this section, (x,y,z), (x′,y′,z′), etc., without subscripts are axis while (x

i

,y

i

,z

i

), (x

i

′,y

i

′,z

i

′), etc., are spot beam coordinates.

1) Rotate the axis (x,y,z) by θ

sat

and ψ

sat

such that the satellite radius (x′ axis) is aligned with the x axis in FIG.

6

. The coordinates of the three spot beam

i

(i=1, 2, 3) centers and the GEM satellite

12

become:

x

i

′=(

x

i

cos(θ

sat

)+

y

i

sin(θ

sat

))cos(ψ

sat

)+

z

i

sin(ψ

sat

);

y

i

′=y

i

cos(θ

sat

)−

x

i

sin(θ

sat

);

z

i

′=z

i

cos(ψ

sat

)−(

x

i

cos(θ

sat

+y

i

sin(θ

sat

))sin(ψ

sat

).

x

sat

′=(

x

sat

cos(θ

sat

)+

y

sat

sin(θ

sat

))cos(ψ

sat

)+

z

sat

sin(ψ

sat

);

y

sat

′=y

sat

cos(θ

sat

)−

x

sat

sin(θ

sat

);

z

sat

′=z

sat

cos(ψ

sat

)−(

x

sat

cos(θ

sat

+y

sat

sin(θ

sat

))sin(ψ

sat

).

Where θ

sat

=a tan(y

sat

/x

sat

) and ψ

sat

=a tan(z

sat

/{square root over (x

sat

2

+L +y

sat

2

+L )}).

2) Find the location of the center of spot beam

i

(i=1,2,3) on the 2-D plane (the satellite does not move during this transformation).

3) Translate the x coordinate of spot beam

i

(i=1,2,3) center and the GEM satellite

12

by 6378 km such that the 2-D plane

100

is now centered at the center of the earth

21

to eliminate the x component:

a) Find the angle between the GEM satellite

12

and the center of spot beam

i

(i=1,2,3):

λ_{{sat}_{i}} = a \sin (\frac{z_{i}^{'}}{d_{1}});

φ_{{sat}_{i}} = a \cos (\frac{d_{2}^{2} + {(R + H)}^{2} - (x_{i}^{′2} + y_{i}^{′2})}{2 d_{2} (R + H)});

With

d_{1} = \sqrt{{(x_{sat}^{'} - x_{i}^{'})}^{2} + {(y_{sat}^{'} - y_{i}^{'})}^{2} + {(z_{sat}^{'} - z_{i}^{'})}^{2}};

d_{2} = \sqrt{{(x_{sat}^{'} - x_{i}^{'})}^{2} + {(y_{sat}^{'} - y_{i}^{'})}^{2} + {(z_{sat}^{'})}^{2}} .

Check if x

i

or y

i

is negative. If yes, φ

sat

i

=−φ

sat

i

;

b) Find the location of the center of spot beam

i

(i=1,2,3) on the 2-D plane

100

. Then, translate the x coordinate of spot beam

i

(i=1,2,3) center and the GEM satellite

12

by 6378 km

x

i

″=0;

y

i

″=H tan(φ

sat

i

);

z

i

″={square root over (H

2

+y

i

+L ″

2

+L )} tan(λ

sat

i

);

x

sat

″=x

sat

′−R;

y

sat

″=y

sat

′;

z

sat

″=z

sat

′.

where R=6378 km; H=35787 km; φ

sati

, λ

sati

, d

1

and d

2

are depicted in FIG.

7

.

4) Translate the coordinates of spot beam

i

(i=1,2,3) center and the GEM satellite

12

such that the center of the 2-D plane

100

concords with the center of the selected spot beam (y

1

, z

1

):

y

i

′″=y

1

″−y

1

″;

z

i

′″=z

1

″−z

1

″;

x

sat

′″=x

sat

″;

y

sat

′″=y

sat

″−y

1

″;

z

sat

′″=z

sat

″−z

1

″.

The 3-D position acquisition problem is now reduced to a 2-D problem with two unknowns y′″ and z′″. On this new plane, the terminal

18

can now search for its location efficiently with the least amount of computations. The terminal

18

will first search on this plane, as shown in

FIG. 8

, along the axis

2

, joining spot beam

1

and spot beam

2

, for a point (y

u2

,z

u2

) with relative power equal to the relative power measured between spot beam

1

and spot beam

2

during the spot beam selection or reselection algorithm. At this point (y

u2

,z

u2

), a line perpendicular to axis

2

will be drawn. The terminal

18

will then repeat the same process between spot beam

1

and spot beam

3

. The terminal

18

will search along axis

3

, joining spot beam

1

and spot beam

3

, for a point (y

u3

,z

u3

) with relative power equal to the relative power measured between spot beam

1

and spot beam

3

during the spot beam selection procedure. Again, at this second point (y

u3

,z

u3

), a line perpendicular to axis

3

will be drawn. The intersection of these two lines will give an estimate of the position of the terminal

18

.

The following algorithm should be used to estimate the position of the terminal

18

once the four steps required to map the three spot beams

22

on this 2-D plane have been completed. As a convention in this algorithm:

The selected spot beam

22

is referred as spot beam,.

The coordinates of the center of spot beam

1

(i=2,3) are denoted y

i

′″ and z

i

″′.

The relative power measured between spot beam

1

and spot beam

1

(i=2,3) through the spot beam selection procedure is referred as max_power_level.

The relative power calculated between spot beam

1

and spot beam

1

(i=2,3) through the position determination algorithm is referred as diff_power

i

.

The coordinates of the point along axis

2

with relative power (diff_power

2

) equivalent to the measured relative power (max

2—

power_level) are denoted y

u2

and z

u2

. Similarly, the coordinates of the point along axis

3

with relative power (diff_power

3

) equivalent to the measured relative power (max

3—

power_level) are denoted y

u3

and z

u3

.

The minimum power level, used to first guess the location of the terminal

18

within the selected spot beam, can vary depending on the satellite inclined orbit. Beam center positions do not change on the ground. The satellite beamwidth is modified to keep the beam center positions fixed on the ground. Consequently, the angle between the center of a beam and the edge of the same beam is not fixed but depends on the satellite position. To calculate the minimum_power_level

2

and minimum_power_level

3

, we first need to calculate the angle between beam

1

and beam

2

and beam

1

and beam

3

respectively. Once we get these angles, we can calculate, via the antenna pattern approximation equation, the powers at these angles.

\begin{matrix} The algorithm shall be perfomed twice for i = 2, and i = 3. \\ 1) Initialize the different variables : \\ • increment = 0.0; \\ • {increment}_{1} = 0.0; \\ • {minimum}_{—} {power}_{—} {level}_{2}, {minimum}_{—} {power}_{—} {level}_{3}; \\ / * Use to first guess terminal location within \\ selected spot beam . See FIG . 8 * / \\ • {maximum}_{—} {power}_{—} level = 0.0 dB; \\ • {diff}_{—} {power}_{2} = {diff}_{—} {power}_{3} = 100.0; / * Temporary \\ value * / \\ • if ({minimum}_{—} {power}_{—} {level}_{i} < {\max_{i}}_{—} {power}_{—} level < \\ {maximum}_{—} {power}_{—} level) \\ y_{ui}^{′″} = y_{i}^{′″} / 2; z_{ui}^{′″} = z_{i}^{′″} / 2; \\ else \\ y_{ui}^{′″} = - y_{i}^{′″} / 2; z_{ui}^{′″} = - z_{i}^{′″} / 2; \\ 2) Perform position determination \\ while (\max_{i—} {power}_{—} level!= {diff}_{—} {power}_{i}) \\ { \\ y_{ui}^{′″} = y_{ui}^{′″} + y_{i}^{′″} * {increment}_{1}; \\ z_{ui}^{′″} = z_{ui}^{′″} + z_{i}^{′″} * {increment}_{1}; \\ H_{fixed} = \sqrt{x_{sat}^{′″2} + {(y_{sat}^{′″} - y_{ui}^{′″})}^{2} + {(z_{sat}^{′″} - z_{ui}^{′″})}^{2}}; \\ H_{1} = \sqrt{x_{sat}^{′″2} + y_{sat}^{′″} + z_{sat}^{′″}}; \\ Δ_{1} = \sqrt{y_{ui}^{′″2} + z_{ui}^{′″2}}; \\ {Δφ}_{1} = a \cos (\frac{H_{1}^{2} + H_{fixed}^{2} - Δ_{1}^{2}}{2 * H_{1} * H_{fixed}}); \\ H_{i} = \sqrt{x_{sat}^{′″2} + {(y_{sat}^{′″} - y_{i}^{′″})}^{2} + {(z_{sat}^{′″} - z_{i}^{′″})}^{2}}; \\ Δ_{i} = \sqrt{{(y_{i}^{′″} - y_{ui}^{′″})}^{2} + {(z_{i}^{′″} - z_{ui}^{′″})}^{2}}; \\ {Δφ}_{i} = a \cos (\frac{H_{i}^{2} + H_{fixed}^{2} - Δ_{i}^{2}}{2 * H_{i} * H_{fixed}}); \\ Δ D_{1} = 10 * \log [{(\frac{\sin (\frac{2 π * ap}{λ} \sin ({Δφ}_{1}))}{\frac{2 π * ap}{λ} \sin ({Δφ}_{1})})}^{2}]; / * ap is the \\ aperture = 4.5 m * / \\ Δ D_{i} = 10 * \log [{(\frac{\sin (\frac{2 π * ap}{λ} \sin ({Δφ}_{i}))}{\frac{2 π * ap}{λ} \sin ({Δφ}_{i})})}^{2}]; / * Power in dB * / \\ {diff}_{—} {power}_{i} = Δ D_{1} - Δ D_{i}; / * Calculated relative \\ power in dB * / \\ if ({diff}_{—} {power}_{i} > \max_{i—} {power}_{—} level) \\ {increment}_{1} = \frac{- 1}{4 * increment}; \\ else \\ {increment}_{1} = \frac{1}{4 * increment}; \\ increment = 2 * increment; \\ } \\ if ((y_{u2}^{′″} ⩵ 0.0) && (y_{u3}^{′″} ⩵ 0.0)) \\ { \\ y_{new}^{′″} = 0.0; \\ z_{new}^{′″} = \frac{(z_{u2}^{′″} + z_{u3}^{′″})}{2.0}; \\ else if ((y_{u2}^{′″} ⩵ y_{u3}^{′″}) && (z_{u2}^{′″} ⩵ z_{u3}^{′″})) \\ { \\ y_{new}^{′″} = \frac{(y_{u2}^{′″} + y_{u3}^{′″})}{2.0}; \\ z_{new}^{′″} = \frac{(z_{u2}^{′″} + z_{u3}^{′″})}{2.0} \\ } \\ else \\ { \\ \begin{matrix} y_{new}^{′″} = (z_{2}^{′″} * y_{3}^{′″} * y_{u3}^{′″} + z_{u3}^{′″} * z_{2}^{′″} * z_{3}^{′″} - \\ y_{2}^{′″} * z_{3}^{′″} * y_{u2}^{′″} - z_{u2}^{′″} * z_{2}^{′″} * z_{3}^{′″}) / \\ (z_{2}^{′″} * y_{3}^{′″} - y_{2}^{′″} * z_{3}^{′″}); \end{matrix} \\ z_{new}^{′″} = ((y_{2}^{′″} / z_{2}^{′″}) * (y_{u2}^{′″} - y_{new}^{′″})) + z_{u2}^{′″}; \\ } \end{matrix}

Where ap=4.5 m is the aperture of the satellite, λ≈0.2 m is the wavelength, and H

1

, H

i

, H

fixed

, Δ

1

, Δ

i

, Δφ

1

, and Δφ

i

are graphically represented in FIG.

9

.

Once the position of the terminal

18

has been determined on the 2-D plane

100

, the coordinates of this location need to be mapped back on the surface of the Earth

21

. The following algorithm describes the three steps required to map back the coordinates of the estimated position of the terminal

18

on the surface of the Earth

21

.

1) Translate the coordinates such that the selected spot beam, spot beams, which was located at the center of the 2-D plane

100

, is returned to its original location on the 2-D plane

100

:

x

new

″=x

1

″+R;

y

new

″=y

new

′″+y

1

″;

z

new

″=y

new

′″+z

1

″;

2) Map the coordinates from the 2-D plane

100

back to the surface of the earth

21

:

x_{new}^{'} = \frac{- B \pm \sqrt{B^{2} - 4 * A * C}}{2 * A} where

A = 1 + {(\frac{y_{new}^{″}}{H})}^{2} * (1 + {(\frac{z_{new}^{″}}{y_{new}^{″}})}^{2});

B = - 2 * (R + H) * {(\frac{y_{new}^{″}}{H})}^{2} * (1 + {(\frac{z_{new}^{″}}{y_{new}^{″}})}^{2});

C = {(R + H)}^{2} * {(\frac{y_{new}^{″}}{H})}^{2} * (1 + {(\frac{z_{new}^{″}}{y_{new}^{″}})}^{2}) - R^{2};

y_{new}^{'} = (R + H - x_{new}^{'}) * (\frac{y_{new}^{″}}{H});

z_{new}^{'} = (\frac{z_{new}^{″}}{y_{new}^{″}}) * y_{new}^{'};

with R=6378 km and H=35787 km.

3) Rotate the coordinates by θ

sat

and ψ

sat

such that the GEM satellite

12

is returned to its original location:

x

new

=((

x

new

′ cos(ψ

sat

)−

z

new

′ sin(ψ

sat

))cos(θ

sat

))−

y

new

′ sin(θ

sat

);

y

new

=y

new

′ cos(θ

sat

)+((

x

new

′ cos(ψ

sat

)−

z

new

′ sin(ψ

sat

))sin(θ

sat

));

z

sat

=z

new

′ cos(ψ

sat

)+

x

new

′ sin(ψ

sat

).

Where θ

sat

=a tan(y

sat

/x

sat

) and ψ

sat

=a tan(z

sat

/{square root over (x

sat

2

+L +y

sat

2

+L )}).

The coordinates of the estimated position of the terminal

18

are x

new

, y

new

and z

new

for this combination. These steps are repeated for one to twelve combinations of three spot beams depending on the number of neighbors in the cluster. The final estimated position of the terminal

18

is the averaged of all the estimated positions (x

new

, y

new

, z

new

) from each combination of three spot beams.

POSITION ESTIMATE METHOD #2

The user terminal

18

calculates a set of points of constant relative power equal to the relative power measured from the three strongest spot beams excluding the selected spot beam. These 3 sets of points form 3 curves with intersecting points, as shown in FIG.

10

. If there is no error introduced, only one point would intersect the three curves at the same time and solve the user terminal position search. However, some error is introduced such as beam pointing error, orbit inclination deviation, fading and slot-to-slot gain fluctuations. The 3 closest intersection points between curve

2

& curve

3

, curve

2

& curve

4

, and curve

3

& curve

4

form a confidence region

104

around the exact user terminal location.

To ease the position acquisition problem, the three strongest spot beams, located on the surface of the Earth

21

, will be rotated and translated in space to end on a 2-D plane with two unknowns y and z instead of the three starting unknowns x, y, z. These transformations will not modify the shape of the spot beams. Consequently, the spot beams will not all have the same dimension on this 2-D plane. The cluster of spot beams will undergo three stages of transformation to end on a 2-D plane centered around the selected spot beam:

1) The spot beams and the satellite coordinates will first be rotated by θ (see

FIG. 11

) such that the center of the selected spot beam is aligned with the x axis:

x′=x

*cosθ+

y

*sinθ;

y′=y

*cosθ−

x

*sinθ;

z′=z;

2) The spot beams and the satellite coordinates are then rotated such that the center of the selected spot beam is traversed by the x axis.

x″=x

′*cosφ+

z

′*sinφ;

y″=y′;

z″=z

′*cosφ−

x

′*sinφ;

3) The spot beams and the satellite coordinates are then translated by (x

1

, y

1

, z

1

) such that the selected spot beam center location (x

1

, y

1

, z

1

) becomes (0, 0, 0).

x′″=x″−x

1

″;

y′″=y″−y

1

″;

z′″=z″−z

1

″.

4) Finally, the cluster lying on the surface of the earth is uncurved to lie on a 2-D plane. This transformation will remove the x component of the coordinates of the center of spot beam

i

(i=2,3,4). In order to keep the original dimensions of the system as much as possible such that the distance between the satellite and the center of spot beam

1

remains constant, I will move the satellite with each spot beam. The y and z coordinates of the spot beams and the satellite are not affected by these transformations.

r_{i} = \sqrt{{(x_{i}^{′″} - x_{1}^{′″})}^{2} + {(y_{i}^{′″} - y_{1}^{′″})}^{2} + {(z_{i}^{′″} - z_{1}^{′″})}^{2}};

ω_{i} = a \cos (1 - 0.5 * {(\frac{r_{i}}{R})}^{2});

x

i

′″=x

i

′″+R

*(1−cosω

i

);

x

sat

i

′″=x

sat

′″+R

*(1−cosω

i

);

where R=6378 km and r

i

and ω

i

are as depicted in FIG.

12

.

By performing these transformations, the position search problem was reduced from a 3-D problem to a 2-D problem with y, and z as unknowns. Once these transformation are completed, the user terminal

18

shall first identify the limits of the region where it shall search for its location to further reduce the amount of computations. The user terminal

18

can then start scanning this region. For each y, and z possible combination within that region, the user terminal

18

calculates the relative power from the three strongest spot beams excluding the selected spot beam. If at a certain location y and z, each relative power matches each corresponding relative power measured by the user terminal, a position estimate is found. The relative power between spot beam

1

and spot beam

2

, spot beam

3

, and spot beam

4

is calculated for a certain user terminal position (y

u

, z

u

) as follows:

H_{fixed} = \sqrt{x_{{sat}_{u}}^{{′′′}^{2}} + {(y_{{sat}_{u}}^{′′′} - y_{u}^{′′′})}^{2} + {(z_{{sat}_{u}}^{′′′} - z_{u}^{′′′})}^{2}};

\begin{matrix} H_{i} = \sqrt{x_{{sat}_{i}}^{′″2} + {(y_{{sat}_{i}}^{′″} - y_{i}^{″})}^{2} + {(z_{{sat}_{i}}^{′″} - z_{i}^{″})}^{2}}; & / * For i = 1, 2, 3, 4 * / \\ Δ_{i} = \sqrt{{(y_{i}^{′″} - y_{u}^{′″})}^{2} + {(z_{i}^{′″} - z_{u}^{′″})}^{2}}; \\ Δφ = a \cos (\frac{H_{i}^{2} + H_{fixed}^{2} - Δ_{i}^{2}}{2 * H_{i} * H_{fixed}}); \\ Δ D_{i} = {(\frac{\frac{\sin (2 π * ap}{λ} \sin ({Δφ}_{i})}{\frac{2 π * ap}{λ} \sin ({Δφ}_{i})})}^{2}; & / * ap = aperature = 4.5 m, \\ λ = wavelength . Power in Watts & * / \\ {diff}_{—} {power}_{i} = \frac{Δ D_{i}}{Δ D_{1}}; & / * for i = 2, 3, 4. Relative \\ power in Watts * / \end{matrix}

where H

1

, H

i

, H

fixed

, Δ

1

, Δ

i

, Δφ

1

and Δφ

i

are graphically represented in FIG.

9

.

If the diff_power

i

is equal to the relative power measured between spot beam

1

and spot beam

i

, this point is kept and will be one of the point forming curve

i

. Once the whole region has been scanned, three curves are obtained. The intersection of these three curves will estimate the user terminal position as shown in FIG.

10

.

The signal power measurements are reported by terminals

18

after obtaining a GPS fix, and the signals are in turn calibrated by the network

10

. Thus the signals are kept equalized, and enhanced accuracy is achieved in the power-measurement based position estimations. If better accuracy is still required, the terminal

18

may incorporate knowledge of last known position, the time elapsed since that position fix, the predicted likelihood of movement since its last position fix, and the statistical expectation of location of the terminal

18

into its position estimate.

In the event that GPS positioning is unavailable (due to obscuration, etc.) this position estimate based upon signal strengths is used in lieu of a GPS position fix.

The position estimate is converted to Latitude/Longitude coordinates by the terminal

18

and inserted into the terminal GPS receiver

19

.

Providing Additional Initialization Information

General Case

A typical GPS receiver calculates its position based on knowledge of four GPS satellite positions, and its distance from each satellite. Four GPS satellites are required due to the need to determine four unknowns from four equations: position in each of three dimensions, and exact time.

The availability of an extremely accurate time reference (i.e., accurate to within about two microseconds) would reduce the number unknowns in the equations, and reduce the required number of GPS satellite acquisitions required for a position fix, from four to three, provided some degradation of fix accuracy is tolerable.

Additional positioning information, beyond a rough estimate, may further reduce the number of required GPS signal acquisitions. The communications network

10

may Contain a database of altitude information. Based upon the rough position estimate, an altitude lookup table, and the history of the terminal, an altitude estimate may be formed which would substitute for one of the equations. Also, the network may be able to determine a distance between the terminal

18

and a known reference, such as a satellite or a ground-based antenna. Such information may further reduce the number of satellite acquisitions necessary to compute a position.

Re-broadcasting a current GPS almanac at a high data rate (block

39

of

FIG. 4

) would also enhance GPS acquisition performance of the terminals

18

.

GEM System

The network

10

predicts the altitude of a terminal

18

based on the geography of its spot beam

22

, its rough position estimate within that spot beam

22

, population distribution within that spot beam

22

, and the location history of the terminal

18

. An altitude estimate reduces the number of GPS satellites

14

required for a position fix from four to three.

Also, the network

10

may determine precise GPS time, and the distance from a terminal

18

to the GEM satellite

12

, which has a known position. This is done in the following manner: after the terminal

18

determines its spot beam

22

, it assumes it is located in the center of the spot beam

22

, and performs time synchronization by adding an offset to the received GEM satellite signal time stamp, based on the distance from the GPS satellite

14

to the center of the terminal's spot beam

22

. The gateway

16

monitors transmissions in both directions and calculates a correction for the terminal's local time, with relation to the real GEM satellite time. The gateway

16

then sends a time-correction offset to the terminal

18

. Based on the time-correction offset, the terminal

18

achieves highly accurate (around 2 μS) synchronization to GPS time. Also based on the offset, the terminal

18

may then compute its exact distance from the GEM satellite

12

.

With this invention, time and altitude can be estimated with enough accuracy to be treated as known parameters, solving two of four positioning equations. Acquisition of 2 GPS satellites

14

should suffice in situations where some degradation of fix accuracy is tolerable. But if the GEM position and distance from the terminal GPS receiver

19

are known to high enough accuracy, one GPS satellite

14

could be replaced in the calculations by the GEM satellite

12

, and GPS position could potentially be done with 1 GPS satellite acquisition.

As an additional feature, the GEM network

10

supports GPS almanac re-broadcast. The GPS almanac, as described in Navstar's ICD-GPS-200, gives approximate orbital descriptions of all GPS satellites

14

. If stored by a GPS receiver, it will be able to achieve optimal performance in conjunction with this invention. If a terminal's GPS receiver has a GPS almanac within a few weeks of age, it may begin a directed satellite signal search even before it receives all of the GEM broadcast data, whereas without an almanac, it would wait for the conclusion of the GEM broadcast before beginning a directed search. Thus, the GEM system re-broadcasts the GPS almanac (obtained through the GPS receiver

20

, located at the gateway

16

) in a background channel, so that a terminal

18

may update its almanac whenever needed.

The present invention has been described with reference to specific examples, which are intended to be illustrative only, and not to be limiting of the invention, as it will be apparent to those of ordinary skill in the art that changes, additions and/or deletions may be made to the disclosed embodiments without departing from the spirit and scope of the invention.

Number	Name	Date	Kind
5412389	Olds	May 1995	A
5841396	Krasner	Nov 1998	A
5874914	Krasner	Feb 1999	A
5907809	Molnar et al.	May 1999	A
5945944	Krasner	Aug 1999	A
6006067	Rudowicz	Dec 1999	A
6052561	Rudowicz et al.	Apr 2000	A
6091716	Gorday et al.	Jul 2000	A
6195555	Dent	Feb 2001	B1
6233451	Noerpel et al.	May 2001	B1

Number	Date	Country
WO 9733382	Sep 1997	WO
WO 985157	Jun 1998	WO

Number	Date	Country
60/109963	Nov 1998	US
60/098686	Sep 1998	US
60/098664	Sep 1998	US

Communication network intialization apparatus and method for fast GPS-based positioning

Information

Patent Number

Date Filed

Date Issued

Inventors

Original Assignees

Examiners

Agents

CPC

US Classifications

Field of Search

US

International Classifications

Abstract

Description

Claims

Parent Case Info

US Referenced Citations (10)

Foreign Referenced Citations (2)

Non-Patent Literature Citations (1)

Provisional Applications (3)