Forum

  • Hey folks,

    Does anyone have an opinion or can recite the relevant clause of the 802.11 specification around the following question:

    Are wireless clients required to process all broadcast/multicast frames received, regardless of whether the SSID/BSSID is the one to which it is associated? Wired VLAN isolation aside, does a STA benefit in any way from being on a separate SSID with regard to broadcast/multicast frames over the RF?

    Thanks for your input,

    --Bruce Johnson

  • Hi Bruce,

    I suppose you are only meaning data frames...

    1. Station only process multicast or broadcast data frames from the associated SSID/BSSID.
    - The station filter data frames by the BSSID Field in the frame. (usually after association, station sets this BSSID in to the chipset - non promiscous mode)
    So basically it does not process other BSSID Frames.

    2. In some cases one BSSID supports multiple VLANs. In that case stations will have to process broadcast frames from all these VLANs.

    I don't think standard has a very clear clause on this..

    Regards,
    Kiran

  • By (Deleted User)

    Does anyone have an opinion or can recite the relevant clause of the 802.11 specification around the following question: Are wireless clients required to process all broadcast/multicast frames received, regardless of whether the SSID/BSSID is the one to which it is associated? Wired VLAN isolation aside, does a STA benefit in any way from being on a separate SSID with regard to broadcast/multicast frames over the RF?


    The 802.11 standard is only concerned with mechanisms that operate at Layer 1 (RF) and the MAC sub-layer of Layer 2. Therefore, there is nothing at all in the standard about VLANs.

    Most vendors support multiple SSIDs (logical name of the WLAN) that can be linked to individual VLANs.

    In older legacy deployments, these multiple SSIDs that were linked to the unique VLANS all shared the same BSSID (MAC identifier of the WLAN) which was the radio card of the access point. This is not the best way because a lot of older client cards got very confused when they saw multiple SSIDs that all had the same BSSID.

    Today, most vendors allow for the creation of multipe "virtual" BSSIDs which can be linked together with the multiple SSIDs to unique and independent VLANs.

  • By (Deleted User)

    I should also add that there is talk that the 802.11v amendment may define & standardize Multiple BSSID and Multiple SSID mechanisms in the future. Currently, any implementations of Multiple BSSID and/or Multiple SSID methods is entirely proprietary and up to the WLAN infrastructure vendor.

  • The way I understand it most vendors choose to not forward broadcasts.

  • Hi billem,

    Can you explain a bit more detail about that.. I just feel that the statement is not correct....

    The way I understand it most vendors choose to not forward broadcasts.


    Regards,
    Kiran

  • Thanks for all the responses,

    As always, there's no better site for these discussions.

    My reason for asking is that medical device manufacturers are pushing us to have to have their own SSID for each of their devices, arguing that they could be DoSed by multicast data frames over the RF if they shared the same SSID with other medical devices. This begged the question about what a STA has to process in terms of multicasts on a non-associated (B)SSID.

    Cisco does use the virtual MAC approach with WLAN AP MAC addressing.

    This begs another question - how many BSSIDs are healthy, since this means a whole new set of management and control frames, correct? Cisco supports 16 SSIDs - is this really something we should consider having at our disposal? In a dense environment this means a lot of BSSIDs in the air (N x APs in the service area). How many is too many in terms of RF pollution?

    I agree that broadcasts are usually proxied by controllers.

    Regards,

    --Bruce Johnson

  • DoSed by multicast frames? A separate SSID for each device? ... you already pointed out the obvious here, the RF medium would be nothing but "pollution".

    I have a hard time believing anyone would willingly configure a WLAN like that.

    ...an SSID for each device. That's a good one... X-D

  • Hey,

    C'mon now. Not an SSID for every device, an SSID for every medical device manufacturer (i.e. one for Philips, one for GE, one for Welch Allyn, etc.). You can't do more than 16.

    I would think someone has done quantitative analysis on the impact of increased SSID deployment and RF utilization (not me, of course).

    What's the largest number of SSIDs people have deployed out there?

    --Bruce

Page 1 of 2