Dear Ali,

Half wavelength dipole antenna radiation is similar to your requirement.

AR  Reddy

Hi Ali,

You can use Bi-directional Antenna ,or any antenna that Back lobe is great compare to main-lobe for ex. you can design a yagi antenna with great  Back-lobe.If you cant, determine freq.,bandwidth,vswr, gain, and radiation patterns ,i can design and make for you at 1day [email protected]

you can combine two antennas into one, back to back. 

Any end-fire array with element spacing half a wavelength will produce two main lobes located at 0 and 180 deg in azimuth and 0 deg in elevation. if you need different elevation angles then you will need to increase the element spacing in order to generate a grating lobe. The main beam and grating lobe positions can be controlled by exciting the radiating elements with a proper linear phase distribution. Hope this helps.

A simple solution could be a linear array of vertical dipoles or monopoles. For < lambda/2 distance between elements it has forward beam at 0-deg and backward at 180-deg main of vertical polarization. Horizontal polarization possible with an array of horizontal dipoles. Beamwidth is controllable by changing the number of elements. Sidelobes can be controlled by applying weights across the array aperture. 

have a look to this publication

J-M Floc’h, A El Sayed Ahmad, Y Kokar

« Dual-Beam Antenna Design for Autonomous Sensor Network Applications »

Hindawi Publishing Corporation International Journal of Antennas and Propagation

Volume 2012, Article ID 289681, 7 pages

Half-Wave Dipole (vertical) will have peaks EIRP/G/T at 0 and 180 as well as anything in between (donuts type pattern).  You did not say anything about gain, or for that matter any other requirement for your LEO type application.  I would use two patch antennas with a simple in-phase beam former (3-5 dB gain) it will be a little better than the dipole but a little more complicated to build (it will reject more also from unwanted directions)

The simplest solution is to use a conventional half-wave dipole as others have already commented. If you need higher gain and your design can be planar, try to design the combination of two back-to-back patch antennas with in-phase feeding.

Hello! Ali,

Much of answers have been given and they are essentially in line with your question.

I resound Roberto V. Gatti, Anatoli B., Jean B. and Andrey S. Andrenko.

The underlining principle to achieve your design objective is to exploit the Back-to-Front lobe ratio characteristics of an antenna.

Two possibilities:

Use a single antenna with adjustable beam formation such that you can obtain equal gain in the back and front lobe at to each other.

Alternatively, design a combination of two back-to-back directional antenna with  equal gains at to each other.

More references can be made from different antenna design publications including "Dual-Beam Antenna Design for Autonomous Sensor Network Applications" (J-M Floc’h, A El Sayed Ahmad, Y Kokar).

thanks all for your responses

Simply, You can combine two antennas at 180 degree.

thanks all for your response.

yes you can use two antennas or a dipole bent back to back :D

Do you really want a single antenna with two lobes or two independent antennas pointing in opposite directions?  From your description of an inter-satellite link, it would sound like you're wanting to create some sort of repeater?  Is this narrow band or broadband?  Two patch antennas, isolated by an appropriate ground (i.e. either side of the satellite) would probably give you the two independent antenna scenario.

Half wave dipole provides Figure 8 shape and it has a both lobes (180 away in a space) are major lobes

A half-wave dipole pattern is omnidirectional and has one single toroidal lobe.  It only appears to have two lobes in the E-plane pattern cut through the toroid, and even then it will have a very wide E-plane beamwidth.  In the H-plane it radiates the same in all directions ("infinite" beamwidth).  

Every dipole has a doughnut shaped emission. Every modification to this requires another dipole, active or passive, but a dipole nevertheless. The simplest setup requires 2 dipoles, and by supplying them with signal at proper phases you may obtain broadside or fire end lobes, and the biggest possible gain/complexity factor.

As everybody pointed out, a simple thin dipole will have such a pattern (figure of eight) in the E-plane but it will be omnidirectional in the H-plane. In H-plane too, a bidirectional wide beam pattern can be obtained by using an in phase  two element array of such antennas, placed half wavelength apart. For better gain i.e. narrower beam along with no side lobe, binomial arrays may be used.

Hi Ali

You will find through the link below, a very nice Java applet that can be used to visualize dipole antenna patterns

• http://www.amanogawa.com/archive/antennaB.html

Many other applets are available as well below

• http://www.amanogawa.com/index.html

Best Regards 

Small loop antenna is the cheapest and simplest solution