WPCr 2BJ Courier#|f,+s Roman BoldHelveticaC&PTimes RomanriptHPLAIIP0.PRSx  @hhhhmX@2P ,6 ZFL#|n, +Apple postscriptHPLAIIP0.PRS&n P7hhhhm&P"S^.=K\\===\g.=.3\\\\\\\\\\33gggQzzpf=Gpfzfpp=3=V\=Q\Q\Q=\\33\3\\\\=G3\\\\QX%Xc=.============\3QQQQQzzQpQpQpQpQ=3=3=3=3\\\\\\\\\\Q\\\\f\QQzQzQzQzQ\pQpQpQpQ\\\\\\=\=\===\G=\p3p=p=p3\\\\\z=z=z=fGfGfGfGp=p=\\\\\pQpQN@.S\=Q\\\\\39\7\7!QQ\==\\ff=Q7tggeeggoo.Ig2[-Kye1pe~e\e~x=\\f\z\=\Q\iwUzpNmń\QQ====ńpsfpfzQsGwQ\Q=3QzffQz\Qpi\p\\sQQzpfppps=\pQpppp==\\\\\\\p=i\Q=x26tHHeadingChapter HeadingJ d  ) I. ׃  Right ParRight-Aligned Paragraph Numbers>a݅@  I.   X(# BibliogrphyBibliography݅X  (# SubheadingSubheading0\ E A.  2 v p k ka a8DocumentgDocument Style StyleXX` `  ` a4DocumentgDocument Style Style . a6DocumentgDocument Style Style GX  a5DocumentgDocument Style Style }X(# 2V  v   a2DocumentgDocument Style Style <o   ?  A.  a7DocumentgDocument Style Style yXX` ` (#` a1Right ParRight-Aligned Paragraph Numbers :`S@ I.  X(# a2Right ParRight-Aligned Paragraph Numbers C @` A. ` ` (#` 2O  /a3DocumentgDocument Style Style B b  ?  1.  a3Right ParRight-Aligned Paragraph NumbersL! ` ` @P 1. ` `  (# a4Right ParRight-Aligned Paragraph NumbersUj` `  @ a. ` (# a5Right ParRight-Aligned Paragraph Numbers_o` `  @h(1)  hh#(#h 2Ia6Right ParRight-Aligned Paragraph Numbersh` `  hh#@$(a) hh#((# a7Right ParRight-Aligned Paragraph NumberspfJ` `  hh#(@*i) (h-(# a8Right ParRight-Aligned Paragraph NumbersyW"3!` `  hh#(-@p/a) -pp2(#p Tech InitInitialize Technical Style. k I. A. 1. a.(1)(a) i) a) 1 .1 .1 .1 .1 .1 .1 .1 Technical2${a1DocumentgDocument Style Style\s0  zN8F I. ׃  a5TechnicalTechnical Document Style)WD (1) . a6TechnicalTechnical Document Style)D (a) . a2TechnicalTechnical Document Style<6  ?  A.   2w/a3TechnicalTechnical Document Style9Wg  2  1.   a4TechnicalTechnical Document Style8bv{ 2  a.   a1TechnicalTechnical Document StyleF!<  ?  I.   a7TechnicalTechnical Document Style(@D i) . 23/b efa8TechnicalTechnical Document Style(D a) . Doc InitInitialize Document Stylez   0*0*0*  I. A. 1. a.(1)(a) i) a) I. 1. A. a.(1)(a) i) a)DocumentgPleadingHeader for Numbered Pleading PaperE!n    X X` hp x (#%'0*,.8135@8:y   O*B ` #&txP7 &P# SUMMARY  YF* x#># &n P7z&P##Xx P7 oXP#Two series of contributions on MAC have been presented to 802.11 which are now characterized  B as:  Yx*B1.XAdaptivelypartitioned periodic frame (APF), and.(#  Y*B2.XAsynchronous sequential access (ASA)(#  x#PThe two plans are quite similar in objectives and asserted functionalities but differ considerably  B in certain aspects of implementation.  x#The case for preference of the asynchronous access is presented. The time organization of a  x#frame structure is unnecessary for a future capacity reservation which in any event only need  x#be known to an intelligent controller. There is a strong possibility that the organization of  x#functions by frame time rather than by transaction/transfer will cause avoidable delays and  x#&decreased reliability from undefined states between time separated but related events. The use  x#4of a slotted fixed length frame inevitably fragments the unused time space resulting in capacity  B loss.  x#This conclusion is believed to be generic with respect to regularly periodic frames with allocated  x#slots and partitions for various functions. For comparison, the design presented by K. Natarajan  x#4was chosen as a representative of the slotted frame class because it is more carefully designed  B and completely described than other like proposals.#.&*&*LL"Bj'#j= '#T = j'#Y'#pY>    `   X* # &n P7z&P##Xx P7 oXP#COMPARISON AND COMMONALITIES OF ASYNCHRONOUS SEQUENTIAL  X*ACCESS AND ADAPTIVELY PARTITIONED PERIODIC FRAME MAC PROPOSALS ď d  B  `   X*P Table of Contents`!(#Page ă  B     B XdOVERVIEW OF COMPARED PLANS p>"(# 1  B XX` ` TABLE I TWO ACCESS METHODS ` p>"(# 1  B   B XdCOMPARATIVE CONSIDERATIONS p>"(# 2  B XX` ` Common Functions ` p>"(# 2  B XX` ` Differences ` p>"(# 2  B XX` ` Normalization ` p>"(# 3  B XX` ` Performance Criteria ` p>"(# 3  B XX` ` Implementation Feasibility ` p>"(# 4  B XX` ` Longevity ` p>"(# 4  B XX` ` ATM Compatibility ` p>"(# 4  B   B XdCONCLUSIONS p>"(# 4  B   B XdReferences p>"(# 5  B ԛ  B   B "0.&*&*LLB"  S* XXXX    ` u # &n P7z&P#COMPARISON AND COMMONALITIES OF ASYNCHRONOUS SEQUENTIAL ACCESS  S*tAND ADAPTIVELYPARTITIONED PERIODIC FRAME MAC PROPOSALS   S*4,(#҇ #&n P7z&P# ` d OVERVIEW OF COMPARED PLANS  xh Two series of contributions on MAC have been  4 presented to 802.11  T* 4 1.XOne of the plans using a regular periodic  4 frame adaptively partitioned between  4 uplink, downlink and random access  4 services was first presented by K. S.  T0 * 4 Natarajan, T* 8&*&* ׍X"Medium access control protocol for wireless LANs (an update)," K. S. Natarajan, IBM Research Division, IEEE 802.11/9239, 3/92(# and later dimensioned and  4 L quantitatively analyzed by R. O.  T * 4 LaMaire. T* &*&* ׍X"Performance of a Reservation MultipleAccess Protocol," R. O. La Maire, IBM Research Division, IEEE 802.11/92108, 9/92(#ƺ This well described plan is  4 taken as representative of a class.  T * 4 2.XA single channel MAC asynchronously  4 but sequentially used for sharing and for  4 avoidance of cochannel interference  Tx*^proposed by C. A. Rypinski T* (&*&* ׍X"Sequentiallyused Common Channel Access Method, C. Rypinski, IEEE 802.11/9195, 8/23/91(#ƚ  T* &*&* ׍X"Architecture and Access Method Analysis for Integrated Voicedata Short Reach Radio Systems," C. Rypinski, WINLab Workshop (Rutgers), 4/28/92(# in 1991."x.&*&*LLl^"  hThe two plans are quite similar in objectives and  hfunctionalities but differ in implementation. Both  hJ plans have in common contention on request,  h exclusivity of channel use for all subsequent  h transfers, division of packets into shorter  hX segments for transmission, and provisions for  hN retransmission of failed segment transfers.  h< While detail changes and improvements have  h| been made in both plans since the cited  h references, the principles involved appear to  h have remained constant. The two plans are  4 compared in Table I below.  hFIt will be asserted that "asynchronous sequential  h access" is a better choice relative to this and all  hother plans using uniformly periodic frames with  4 regular time slots.H.&*&*LL&4$P v &*&*LLHԯ 4   T*4 0 TABLE I TWO ACCESS METHODS ׃  9pyxdddy  T* xh (#B(#҇In the " adaptivelypartitioned periodic frame  T* xhJ (APF) " MAC functions are sorted by time as  xh defined by frame structure including slots for  xh requests, data inbound, data outbound and  xhsubframe headers. The steps of one transaction  xh appear at assigned positions within the defined  xh frame. These time positions are a secondary  xhaddressing system used in the header messages  xh allocating them for use. There is contention  4 possible on requests but not on subsequent steps.  xh This MAC is first applied to a frequency  xh< hopping PHY where any one hopping pattern  xh2 appears as a single clear channel. Current  xh descriptions consider each hop pattern as an  xh2 autonomous and independent system No  xh provision of this MAC so far deals with  4 interworking on different channels.  xh The maximum transfer rate is 1/Nth of the  xhj possible transfer rate in the total allocation  xht where N is the number of hopping frequencies  4 (N = 75 typical).XK%.&*&*LLl$43'#X  T* hIn the " asynchronous sequential access (ASA) "  h plan there is one radio channel only operated at  h the highest rate feasible within the allocated  h` bandwidth. MAC function grouping is  hX immediately sequential for each transaction or  h transfer (in other MACs described as a 5step h handshake). Each new transaction can begin as  h soon as the preceding transaction is completed.  h8There is contention possible on requests but not  4 on subsequent steps.  hJ The frequency reuse problem is addressed by  h@ nonsimultaneous operation of potentially  h interfering stations/access points within one  h< cluster. The station is not required to change  h@ channels or patterns in any circumstance.  h Capacity may be moved adaptively between  4 sites.  hX The peak transfer rate is that of the allocation.  h The maximum average rate of one accesspoint  hD is 1/Nth of the peak rate where N is the  4 frequency reuse factor (N = 4 assumed).yPddd yJ%.&*&*LLl$4$#&,@,&*&*LLe'#P/*P ԯ &pyx+ddd| y&.&*&*LL%p'#P/*P +'#++|   S*4B(#(#҇X` hp x (#%'0*,.8135@8: